Pumped storage: underground challenges

2 July 2014



As Europe’s push for wind and solar drives pumped storage, part of the design and maintenance challenge for hydro lies underground. Report by Patrick Reynolds


Tunnels are a major feature of many pumped storage schemes, and their design and maintenance needs are key parts of research underway in Norway. The work is vital if the country's hydropower sector is to provide a macro-solution, comprising a large network of pumped storage schemes, to help counter the imbalancing effects of intermittent renewables on Europe's transmission grids, primarily in Germany.

However, while Germany radically shifts its energy mix from nuclear and brown coal towards renewables - especially wind and solar, it is also taking a fresh look at pumped storage, albeit in the form of large individual projects as conventionally done across the world. Potential sites have been identified for studies and plans, including in the states of Thuringia, and Baden-Wurttemberg.

Norway - operational changes

Norway is looking to the potential massive deployment and expansion of much of its southern hydropower assets in the proposed "Green Battery" scheme to help balance fluctuating renewables outputs in mainland Europe. Without such large-scale dynamic and storage resources there is risk to the efficiency and stability of the grids, and also no way of usefully soaking up excess power generation.

Preparation for Norway's possible shift to region-wide, integrated pumped storage management includes due diligence on how existing assets, including tunnels and power plants, would need to be operated. But there is also a need to examine if plants and infrastructure might be negatively affected by the new conditions, and, if so, to address how to mitigate and overcome such potential risk.

Researchers at the Norwegian University of Science and Technology (NTNU) have been examining some of the potential consequences from changes in the operational regimes in Norwegian hydropower plants. The research includes investigating particular arrangements of sediment handling, and its presence, in hydropower tunnels, and also the impact and role of its specialty air-cushioned surge chambers.

The studies come under the HydroPeak research programme, which is managed by the Centre for Environmental Design of Renewable Energy (Cedren).

In March, NTNU Professor Anund Killingtveit led a small team to brief the latest annual national hydropower industry conference - Produksjonsteknisk konferanse (PTK) - on the studies and work done to date. The focus was partly on Hydropeak's Work Package 6 - 'Load Fluctuation: Tunnels'. The team gave a related briefing on aspects of their work to IWP&DC.

Killingtveit says: "The introduction of intermittent and more variable renewable power plants (wind, solar and small hydro) lead to more unstable generation, and a need for balancing from other sources."Norway, he says, 'with practically no thermal generation,' relies on reservoir hydropower plants to do the balancing, resulting in variability in generation. The trend is expected to continue, and has the potential to increase notably, not least due to the Green Battery initiative.

“Rapid changes and more frequent start/stop will become the normal operating regime for many Norwegian hydropower plants," he adds.

Changes in risk at Norwegian hydro plants

"Introducing different operational regimes...might impact the structure of the underground infrastructure, and the durability of turbines"

Nearly all Norwegian plants of more than 10MW installed capacity have power tunnels, and most of those are unlined, except for near underground powerhouse.

Introducing different operational regimes, and the consequent effects for flow dynamics in the tunnels, might impact the structure of the underground infrastructure, and the durability of turbines, says Killingtveit. It might also cause some local effects in rivers or other water bodies into which the plants discharge, he adds.

Such potential consequences need also to be weighed when assessing refurbishment or uprating projects, he notes.

“It is important to know what kind of impacts the more variable flow could have," says Killingtveit, "and how large are the changes that can be tolerated without creating problems."

The kinds of changes that could arise and, therefore, need to be investigated, he says, include alterations in the risk of erosion within the tunnels, and the sediment load carried through turbines. Also, there could be ice or flow variability problems for fisheries where the tailrace discharges.

Potential for damage, hold-ups and costs have been shown at some existing plants already where some changes occurred. In those cases, he says, increased transport of sand, gravel and even small rocks at plants 'led to serious damage and long stops for repair.'

Sediment in tunnels

Norway's major expansion in hydropower took place over the 1960s-80s, and the projects involved significant underground works. A standard procedure in construction was to leave a layer of spoil on the tunnel floor, explains Killingtveit.

"Sand traps are a key element in helping to lessen the sediment loads in water flowing along the tunnels"

The layer was approximately 150mm thick and was available to be employed later for maintenance, if needed, as a local resource to help carry out repair work.

“As long as the flow was fairly stable, and velocity not larger than anticipated, this layer was usually stable," says Killingtveit. Only some fines would be carried towards the turbines, and there were not too many problems as a result, generally, he adds.

“But with increased capacity or increased use of the power plant for peaking, this may change," as the tunnel bed layer might be disrupted, he says.

Sand traps are a key element in helping to lessen the sediment loads in water flowing along the tunnels. They have been located upstream of penstocks, and were created by locally enlarging the tunnels which causes the flow velocities to reduce and the heavier material to sink to the bed. At intervals, the deposits are cleared as part of the maintenance regime.

Difficulties can arise, though, should the traps have been under-designed, or not working according to the design, especially for increasing flow velocities, says Killingtveit. In such circumstances, even small stones may not always sink out of the flow but can be carried into the plant and damage the turbines, he notes.

PhD student Kari Bråtveit is working on developing tools to predict the effects of load fluctuation in hydraulic systems, including effects on sedimentation and sand traps.

Surge chambers

Surge chambers are key elements for managing the flow dynamics of hydraulic tunnels; they dampen the impacts of flow and pressure changes, especially those induced by sudden starts and stops of turbines - or pump-turbines.

Open air surge chambers are common in hydropower, and are usually constructed at the junction of upper headraces (lower pressure) and the steeper tunnels and steel-lined penstocks (higher pressure), such as in many Alpine projects. However, a different concept was developed in Norway due to problems with high overburden, which rendered open air surge chambers impossible.

Norway's excellent geology, with its beneficial rock quality and rock stress conditions, has allowed construction of closed surge chambers entirely underground. The trapped air in the chambers acts to cushion, and dampen, the surge flows in the tunnel. The air-cushioned surge chambers allow for the power tunnels to be constructed differently - as a continuous, pressurised sloping tunnel from the upper to the lower reservoir. Also, they can be located close to - in fact, just upstream of - underground power plants.

As changes in operational regimes of power plants in Norway would introduce more flow and pressure variability into the hydraulic systems, and the potential consequences for their structures, such as air-cushioned chambers, need to be examined. NTNU research on surge chambers under HydroPeak is being undertaken by PhD student Kaspar Vereide.

Surge Chamber Research

The surge chamber is the critical component for controlling hydraulic transients (water hammer and mass oscillation) in the tunnels, says NTNU PhD student Kaspar Vereide.

At PTK 2014, he covered the need, the types (open air, air-cushioned, vertical and inclined) and sizing of surge chambers before reviewing new developments, including understanding the impact of "hydropeaking" regimes on the structures.

Vereide explains that his research was originally scoped to look at long waterways for pumped storage, but has come to focus much more on surge chambers.

"“Pumped storage plants have massive amounts of hydraulic transients compared to regular power plants, and the surge chamber is therefore of crucial importance"

“Pumped storage plants have massive amounts of hydraulic transients compared to regular power plants, and the surge chamber is therefore of crucial importance," he says.

His work has included measurements for numerical modelling of a number of plant waterways, including those of the Oksla, Jukla, Duge and Tonstad plants in Norway.

The current stage of his research work is, unusually, the physical modelling of an entire hydraulic system of a plant - the Torpa scheme, including its air-cushioned surge chamber. The Torpa plant is owned by Oppland Energi AS through its subsidiary Eidsiva Vannkraft AS. It was built in 1989 near Lillehammer, and has a head of 470m.

Vereide says the model is the first time the complete waterway of a Norwegian hydro system has been constructed - the upper and lower reservoirs, the headrace tunnel, penstock, turbine, tailrace tunnel and the air-cushioned surge chamber. NTNU supervisor on the research project is Prof Leif Lia.

Assembly of the physical model began late last year. The model has been constructed at a scale of 1:68.6, is 150m long and 12m high, effectively running the length and back of the hydraulic laboratory, and taking up more than the floor to ceiling height.

Preliminary tests have been underway to verify the model design, and its ability to correctly simulate the physics of the real waterway system. Due to the difficulty of scaling the physics, the model is using under-pressure for the tests, which are to get underway soon and run until mid-2015.

Vereide says that in addition to project-specific results, it is hoped the findings will generate more general conclusions that are applicable to the underground waterways, and air-cushioned surge chambers, of the wider Norwegian hydropower sector.

The air-cushioned surge chambers are still believed to be state-of-the-art for reasons of being less expensive, and offering both better dampening and regulation capabilities, and reduced environmental impact, says Vereide.

He adds, though, that the research work also includes testing the use of throttles at the air-cushioned surge chamber, which has not been done before.

Monitoring & modelling

The hydrodynamic properties of underground systems - especially those with long tunnels, and including intakes, surge chambers, penstocks, gates and valves - 'can become very complex,' says Killingtveit.

"Numerical simulation can help establish how much change might be acceptable under a new operating regime"

Flow velocities might arise that could cause erosion or 'even flushing out of sediments in the sand trap.'

Killingtveit notes that such critical events can be difficult to predict, observe and understand, not least because tunnels do not usually hold enough instrumentation to provide sufficient monitoring data. He says an approach to overcoming the problem is to use numerical modelling, so long as there are sufficient details - spatially, and also in time behaviours related to different operational requirements of the transmission grid.

The numerical simulation can help establish how much change might be acceptable under a new operating regime, but also help to assess the measures that might be taken to counter potential problems, he says.

In the research on sedimentation, field work at plants like Tonstad has included measuring the insitu dewatered tunnels by laser scanning. The data are then used to help construct 3D models for numerical analysis of flow in the hydraulic tunnels, says Bråtveit. Physical modelling work in the laboratory is also being undertaken, she says.

Numerical modelling also features strongly in modern surge chamber design, notes Vereide. While physical modelling was not the primary choice, mainly due to size and scaling difficulties, the simple equations used in early designs gradually became more detailed and complex, and have given way to numerical modelling. Yet, as the surge tanks get more complex, there is a verification role for physical modelling, he notes.

Vereide adds that the combination of physical and numerical modelling has become standard for large pumped storage projects in Europe - and when they 'are to be constructed in Norway, we expect it to be the same.'

Germany - project studies

While Germany looks at the large-scale, strategic offering of pumped storage support from Norway, the possibility of constructing some new domestic capacity is under examination. Studies have been underway on various potential pumped storage projects, including: a choice of two sites for Strabag in Thuringia; and, the upgrade and expansion of EnBW Kraftwerke's Rudolf Fettweis scheme at Forbach, in the Baden-Wurttemberg, with planning and design by a joint venture of Lahmeyer group with Geoconsult.

A little over a year ago, Strabag announced it was planning - along with future investors - to build a pumped storage project in central Germany. The candidate sites for the project are in Ellrich, and Leutenberg/Probstzella, and studies are still continuing, says a spokeswoman for the construction group.

Both prospective projects call for underground infrastructure to link their respective upper and lower reservoirs, and house the power plant complex. At the concept stage, the 640MW Ellrich scheme was envisaged to require a reservoir capacity of 6.3Mm3; and, the 380MW Leutenberg/Probstzella project would need a 4.1Mm3 reservoir.

In its announcement, Strabag said at the time the estimated investments to build either power station were more than Euro750 million and Euro450 million, respectively. The company envisaged a project being commissioned by the early 2020s.

In south west Germany, EnBW's complex at Forbach comprises a clutch of small plants, except for the 43MW Schwarzenbach hydropower station near the dam of the same name. The scheme was constructed in two stages almost a century ago.

EnBW appointed Hydroprojekt Ingenieur, a Lahmeyer subsidiary, along with the parent group and also Geoconsult, to look in detail at the new scheme to build two pumped storage plants - a 200MW new facility, and a 70MW plant to replace the existing Schwarzenbach power station. The new scheme was conceived by Hydroprojekt Ingenieur and Lahmeyer in early studies between 2007-10.

Both of the proposed new pumped storage plants would have underground infrastructure - the larger scheme having tunnels and a powerhouse cavern complex; and, the smaller scheme would see the plant housed in a shaft and the lower reservoir underground.

Pumped storage: the resurgence

Pumped storage is resurging, thanks to intermittent renewables and the needs of energy storage.

Norway can offer a macro solution of networked pumped storage schemes to Germany and Europe, and Germany itself is also exploring possibilities for more local project contributions.

But whatever type of schemes, designs and locations are chosen to help balance the myriad effects of growing intermittent renewables on the transmission grids of the Continent, it looks certain that underground construction, maintenance and research will be a fixed part of the pumped storage picture.


Author notes

Patrick Reynolds is associate editor at IWP&DC

Large sand trap at Tonstad plant, Norway. Courtesy NTNU
Scanning the sand trap at Tonstad plant, Norway. Courtesy NTNU
Model element of air-cushioned surge chamber. Courtesy of NTNU
Schematic of Torpa plant physical model at NTNU. Courtesy of NTNU


Privacy Policy
We have updated our privacy policy. In the latest update it explains what cookies are and how we use them on our site. To learn more about cookies and their benefits, please view our privacy policy. Please be aware that parts of this site will not function correctly if you disable cookies. By continuing to use this site, you consent to our use of cookies in accordance with our privacy policy unless you have disabled them.