Pumped storage: a new project for Wales

A new pumped storage scheme for emergency electricity generation is to take shape in North Wales’s Snowdonia National Park. The SPH development team explain the thinking and the engineering behind the project.

With the highest mountain in Wales as its centerpiece, Snowdonia’s natural beauty is a magnet for some six million visitors a year. The same glaciated geology that helps make Snowdonia so valued as a wild place also makes it attractive as a location for grid-scale electricity storage.

Dinorwig, currently Britain’s largest pumped storage scheme at 1.7GW, was built there on the northern shore of Llyn Padarn near the town of Llanberis. But that was 30 years ago. It seems astonishing to many involved in the energy sector that despite Britain’s dash for wind and solar, with all that their intermittency implies for grid instability and wasteful constraint arrangements, that three decades should have passed without the creation of further grid-scale storage. Yet that is indeed Britain’s record. While other countries were building storage to mitigate the grid instability caused by intermittent renewables such as wind and solar, successive British governments have been unaccountably blind to this need.

Today in Britain a coalition of energy academics and industry figures including two former chief scientific advisors, National Grid, DNOs, renewables companies and green NGOs such as Friends of the Earth are all saying that Britain has insufficient storage. Meanwhile the government has failed to show that it either understands the immediacy of the problems caused by insufficient storage, or that it has any clue how much new storage will be required as renewables increase through 2020, 2025 and onwards.

In contrast, Portugal, one of a number of countries whose governments have shown more foresight, is planning to deploy 1GW of new storage for every 3.5GW of renewables. By the Portuguese formula Britain is currently 2GW short of the storage it needs and will be more than 6GW short by 2020. With insufficient ability to store excess production by renewables when demand is low, then release it when demand is high or the wind ceases to blow, Britain’s grid is needlessly volatile and exposed to the risk of blackouts.

Energy bill payers might justifiably expect the UK to have a less wasteful, more joined-up energy strategy, but the government’s sole response so far has been to pay wind farms to shut down when winds are high, and pay owners of previously mothballed thermal power stations to hold them ready for action when the wind drops.


Outside the UK, pumped hydro storage – as harnessed by Dinorwig – remains the dominant technology. Used to provide more than 90 per cent of the world’s storage needs, it is time-proven and able to scale to a size large enough to be of use in balancing country-wide electricity grids. In the UK though, there has been an assumption that no sites remain in the British Isles where further pumped hydro could be built.

QBC, a Cambridge-based storage developer is proving that assumption wrong. It has carried out a UK-wide GIS-based survey and found suitable sites for some 15GW of pumped hydro. Not all locations are what may be thought of as conventional. Some will re-purpose brownfield land while others will use coastal features and sea water. Others still will use drinking water.

The first of these schemes is due to begin operating by 2019. Sited on the southern shore of the same lake that neighbours Dinorwig, the 600MWh Glyn Rhonwy scheme is using disused slate quarries – rather than exploiting natural geographical features such as the cwm that Dinorwig uses as its top reservoir. QBC has formed a subsidiary company, Snowdonia Pumped Hydro, to take the scheme through to operation. With planning permission granted in late 2013 and ground investigations now underway, construction companies could be invited to tender for the build before the end of 2015.

Glyn Rhonwy Scheme – The Nuts and Bolts

The site of the Glyn Rhonwy Pumped Storage facility is on the slopes of Cefn-Du, a hill that rises from the south bank of Llyn Padarn. Cefn-du was used by Marconi in his experiments with long distance radio communication and was from where in 1918 the first radio transmission from the UK to Australia was made.

Open pit slate mining had begun on its northern face overlooking the lake in around 1840, but by 1930 all slate quarrying had ceased. Ninety years of dangerous and backbreaking work by the men of the area had left the hillside scarred by multiple quarries and slate tips.

Llanberis 2

Used during World War Two as a site for explosives storage and disposal, it was subsequently handed by the British Ministry of Defence to Gwynedd Council. In 1998 the council began leveling the lower ground and building access roads to create a site suitable for light industry. No applicants for the building plots came forward and there followed a decade of inactivity punctuated by ultimately fruitless development proposals.

QBC approached the council with the idea for a pumped storage development whose primary features were a repurposing of the two largest quarries into storage lagoons, and the construction of an associated hydro power station on part of the lower ground of the site. Council members on the specially formed Glyn Rhonwy Working Group saw that the proposal would not only create valuable full time jobs for the town of Llanberis, but also provide a long-lived source of rates revenues as well as potentially act as a catalyst for the long-hoped for occupation of the industrial site.

QBC was invited to design a scheme and chose to work in partnership with civil engineering company Aecom whose track record includes the design of hydroelectric schemes and power stations around the world. Cadw, the Welsh government’s guardian of historic sites, Natural Resources Wales (formerly the Environment Agency Wales) and Gwynedd Council itself were involved closely in the process with the aim of achieving a visually unobtrusive design that met with maximum approval from key stakeholders.

The topmost quarry, Chwarel Fawr, like its lower counterpart Glyn Rhonwy, is typically steep sided. One of the earliest construction tasks will therefore be to reprofile the quarry sides, using rock anchors where necessary, to make them stable and able to withstand fluctuating water levels. Spoil from this process and elsewhere on the site will be used to create 20m high dams at the lip of each reservoir. The dams will be lined to make them impermeable and founded in the bedrock so that a grout curtain can be injected to prevent seepage underneath them.

The finished reservoirs will each hold around one million cubic metres of water, with the topmost having a water depth of 60m. The height differential between lower and upper lagoons is 300m, requiring a penstock 1.5km long to transfer the water between the two. This is to be tunneled through the rock of Cefn-Du whose slopes are not regular. The tunnel’s effective depth below ground level will therefore fluctuate between 50 to 80m, plunging to 100m below ground level before entering the turbines. The 4m diameter tunnel will be steel lined towards the bottom of its run where pressures will be highest.

The turbine house will be located on part of the lower ground that has been set aside by the council for light industry. It will be a shed-like building, faced in local slate to match other local buildings. It will stand over a 120m deep, 30m diameter concrete shaft, near the bottom of which will be sited the turbines and their associated generator sets. Spoil from the penstock tunnel boring, and the excavation of the turbine hall shaft, will mostly be recycled into the dam structures, with any excess tipped within the curtilage of the site. At the top pond location excess material will be used to form heaps mirroring those that are already a feature of the surrounding post-industrial landscape. In this way the need to remove any material from the site by road will be avoided.

An overflow will link the lower dam to Llyn Padarn, as will a scour valve to allow the water level at the facility to be lowered if necessary. A small discrete pumping station will be buried adjacent to Llyn Padarn to enable the facility to be filled with water abstracted under license from the lake. Filling is expected to take around 12 months and will begin as soon as the lower reservoir has been completed.

Because the upper reservoir is being formed from an existing quarry that is narrow and deep, rather than being the ideal shallow and wide saucer shape that would be excavated on a greenfield build, the head will reduce rapidly as the lagoon empties. Water pressure at the turbine house will therefore tail off over the six-hour run-time of the facility, requiring the flow to be throttled in order to deliver a stable output.

Another notable feature is that there will be no surge pond or surge shaft to mitigate against equipment damage by negative pressure waves. Instead, the turbines buried some 80m below ground level will be at sufficient depth to create a pre-pressurised system in which return pressure waves never reach such a low level to cause equipment damage.

The turbines for the scheme will be buried some 80m underground

The two 49.9MW turbines will be reversible Francis configuration with single impellers, able to both pump and generate. Encased in mass concrete they will be configured so that one can be maintained while its partner is operating. The developer has requested that the District Network Operator connects the facility to the local grid 8km away at Pentir by buried cable, rather than overhead lines.

Ian Gillies, Aecom’s project director for hydro business in UK and Europe, said that the design of the scheme had been testing but satisfying. “Everyone involved is very excited about getting this project away. It’s a proper engineering challenge – not only the sort of scheme hasn’t been built in Britain for some considerable time, but requiring innovation too. The reprofiling of the quarries, low variable head operation and the pre-pressurising of the turbines are all proper challenges. As an engineer this is the sort of project you really want to get stuck into.”


Pumped Storage – The Commercial Perspective

Pumped Storage charges up on off-peak electricity, but is not 100 per cent efficient at recovering the electricity used. A modern plant can expect to achieve a cycle efficiency of 80 per cent; meaning for every five units of energy purchased, four can be returned for sale. The breakeven point can be calculated by taking the cost of off-peak electricity and applying the efficiency loss and operational costs.

Because the breakeven price tracks the cost of electricity, and because the price of electricity tracks the price of fossil fuel, pumped storage automatically hedges against any peaks or troughs in fuel prices. Inflation is also correlated to energy prices, so pumped storage also hedges against inflation risks.

With wholesale gas prices currently around £22/MWh a modern Open Cycle Gas Turbine (OCGT) has a breakeven price (before OpEx) of £55/MWh. Combined Cycle Gas Turbines (CCGT) are much slower reacting than OCGT, but are more efficient, yielding a breakeven price (before OpEx) of £38/MWh.

Despite the absence of substantive government encouragement for pumped storage the Glyn Rhonwy scheme will have a comparable breakeven price to gas generation fleet that is slower to start up and shut down, and less reliable too. In other words it will be more useful as a grid balancing tool than a gas power station, but cost about the same.

The numbers stack up favourably for Glyn Rhonwy because the re-purposing of existing quarries means the cost of major excavation and spoil removal works is avoided. At a development cost of £160m, and with an output of 600MWh, investment payback will be achieved within 15 to 20 years, while carbon payback will be achieved within just six to nine months.

Grid-scale storage like that at Glyn Rhowny will become increasingly valuable to Britain because of the volatility of renewables. Wind often blows hard at times of low demand – often at night. National Grid is already paying the operators of Britain’s existing pumped storage sites to soak up excess power at such times. The frequency of these events, and of minute-by-minute fluctuations during daylight hours, will only increase as the percentage of renewables continues to grow – increasing the need for the services that storage provides.

SPH anticipates that the Glyn Rhonwy facility will operate a hybrid commercial model combining simple day-night arbitrage with active trading. Glyn Rhonwy will be buying electricity – or being paid to absorb it – at nighttime when it is generally cheaper, then regenerating it for sale at a higher price during the morning and evening peaks. It will also sell short bursts of power, and short duration absorption services into the balancing mechanism, the tool used by National Grid to help it balance supply against demand.

With the economic balance surrounding pumped storage becoming inexorably more favourable over the coming 15 years renewables companies will likely look to add storage as a complement to their windmill or solar fleets. This will enable them to increase the accuracy of their output by soaking up peaks in generation and smoothing out troughs. Pumped storage used in this way to leverage a generator’s other assets can be expected to improve returns across a whole generation portfolio.