A few weeks ago I attended a small, commercial, energy storage conference in Brussels organised by Energywise where I heard a most intriguing talk on building a large pumped storage hydro scheme in Holland. The talk was delivered by Dr Jan Huynen, the president of SOGECOM who struck me as being a very serious energy engineer. The project is nearing fruition, with a €1.8 billion price tag and 1.4 GW of supply for 6 hours yielding 8 GWh per daily cycle, this is no toy. Holland is of course totally flat!
Is this just another Green pipe dream? Or does it offer a solution to the apparently intractable problem of energy storage? There is of course nothing new about pumped storage hydro. But all existing schemes use natural relief and elevation to create the head required to store gravitational potential energy that creates pressure and power. What makes Flat-land Large-scale Electricity Storage (FLES) unique is that the whole system is located underground (Figure 1). This of course adds cost but also, as we shall see, it offers substantial benefits.
Figure 1 FLES employs a small surface reservoir and deep lower reservoir with all tunnels, pipes and generating kit below the surface .
Why Energy Storage?
Electricity demand in advanced societies follows a very specific and precise pattern that mirrors human and societal behaviour. Demand is always higher during the day than at night when most people are asleep. It is higher in winter when we spend more time indoors keeping warm. It is higher during the week when the wheels of commerce are more active. (Figure 2) The demand pattern may vary from country to country.
Figure 2 The very specific cyclical pattern of electricity demand reflects individual and societal adaptation to using the energy stores of fossil fuels and uranium in electricity production. Using energy stores has resulted in a very well ordered society that is not equipped to use the intermittent energy flows from renewable energy sources.
Society has evolved in this way because we had access to energy stores, namely fossil fuels and uranium. We can release energy from those stores at a rate that matches demand exactly.
It has now been decreed by the majority of OECD governments and organisations that we must abandon this system upon which our civilisation was built to use instead stochastic energy flows from renewable energy sources. This decree is based upon multiple layers of deception. But whether we like it or not, we now have significant and growing amounts of renewable energy flows on our electricity grids that threaten their integrity and our prosperity. The only way to convert third class energy flows into first class energy stores is to somehow store renewable electricity for use when we want to use it. As we have seen in recent posts, the scale of this challenge is formidable [2, 3, 4, 5, 6].
Without delving into the details there is only one storage show in town and that is pumped storage hydro. This technology alone provides the energy efficiency, storage capacity and power delivery to provide a meaningful and economically viable solution (Figure 3).
Figure 3 The chart compares the Power and Storage of batteries, compressed air / thermal and pumped storage. Note the Log – Log scale. Individual pumped storage schemes are 10 times bigger and are more energy efficient than the nearest compressed air rivals. Chart from  originally from RWE.
Historically, pumped hydro in Europe was built primarily to store surplus nuclear power at night and to release this surplus into the daytime peak demand. The business model was founded on price differentials between day and night time electricity prices. However, despite what appears to be clear social benefits, pumped storage in Europe remains an insignificant fraction of total generation :
UK 3 GW
France 4.5 GW
Germany 6 GW
Austria 8 GW
Total pumped storage = 21.5GW compared with 423GW of installed generating capacity in those 4 countries (Austria 23 France 129 Germany 177 UK 94).
High capital cost and insufficient pay back is likely to be one obstacle to larger scale deployment of pumped storage. But environmental harm, public protest and a paucity of suitable sites have likely also played a role. Traditionally, pumped storage was the domain of mountainous countries and largely flat countries like Holland, Denmark and England were automatically excluded.
The Dinorwig pumped storage hydro scheme in N Wales was built into a mountain (Figures 4, 5), that once was a slate quarry, in order to minimise the environmental impact upon the Snowdonia National Park. Construction began in 1974 and took 10 years to complete. Dinorwig differs from FLES in that the upper and lower reservoirs are on the surface and the scheme is therefore constrained by surface relief.
Figure 4 The Dinorwig pumped storage scheme in Wales comprises two surface reservoirs offset by about 500 m relief.
Figure 5 The Dinworig inlet manifold provides testimony to large scale underground engineering works over 30 years ago.
Some vital statistics
Machine hall volume 211,140 m^3
Tunnels 16 kms
Reservoir volume 6.7 million m^3
Head 542-494 m
Power 1.8 GW
Energy stored 9.1GWh
Duration 5 hours
Cost £425 million
Flat-land Large-scale Electricity Storage
This is where Flat-land Large-scale Electricity Storage comes in. The concept is a very simple modification of technologies that have existed for decades. The upper reservoir may be located on flat land, almost anywhere. The lower reservoir is located below ground in a large excavated cavern. The machine hall that contains the turbines and pumps is located deep underground close to the lower reservoir (Figure 1). The concept offers three significant advantages over conventional pumped storage in that:
1) The head is not limited by natural relief but by the depth to which one may wish to mine that in turn may be controlled by geology. Hence energy storage capacity can be simply increased by digging deeper.
2) The surface expression on flat land would be a reservoir perhaps 400*400m surrounded by a grassy dyke (Figure 6). On flat agricultural land, forest, heath or wasteland the visual impact will be minimal. While the facility will have no recreational value, it will preserve the recreational value of mountainous areas.
3) Given a sound economic model, a scheme may be replicated many times over to provide scale
Figure 6 Artists impression of O-PAC on flat land near Maastricht, Holland. There is little to betray that fact that this is a 1.4 GW power station .
SOGECOM have identified and acquired a site near Maastricht in Holland to build the world’s first FLES facility christened O-PAC. In Holland, a land of soft sedimentary rocks, finding suitable strata for the lower reservoir presented a challenge solved by the identification of a thick (700m), hard, Carboniferous limestone bed at 1400 m depth.
The vital statistics are impressive.
Upper reservoir 400*400*16m
Water volume 2.5 million m^3
Power 1.4 GW
Duration 6 hours
Energy stored 8.4 GWh
Cost €1.8 billion
Life of circa 100 years
The specifications of O-PAC are very similar to Dinorwig, the main difference being the smaller reservoir volume that is compensated by greater head. The main construction difference lies in the need to excavate a lower reservoir at depth that is 10 times larger than the machine hall excavation at Dinorwig. This obviously will add significant cost.
The business model for O-PAC relies upon the diurnal price cycle where off-peak electricity is purchased at night and used to pump water that is released into the day time peak in demand. This price differential is quite significant, off peak below €30 / MWh and peak above €50 / MWh in Holland (Figure 7). However, as with all energy storage systems, O-PAC is not 100% efficient. With an estimated 80% round trip efficiency, 20% more electricity needs to be bought than can be sold.
O-PAC is designed for fast system response and is intended to respond to short time scale (15 minute) fluctuations in the electricity market. Going forward, SOGECOM anticipate that higher penetration of renewables will create greater volatility, and with that, greater market opportunity. Hence the primary role of O-PAC will be to facilitate load balancing. Detailed modelling of the present and future market  implies 250 full cycles producing 2TWh of electricity / year. A back of envelope calculation based on €50 / MWh implies annual electricity sales of the order €100 million.
Figure 7 Dutch electricity demand, black upper curve, left hand scale, MW. Electricity prices, grey lower curve, right hand scale, € / MWh. Mid week off peak prices are often below €30 / MWh while mid week peak prices are typically above €50 / MWh 
I have been puzzled by the status of UK and European energy storage. Politicians and renewables enthusiasts proclaim that intermittency will be solved by storage. And yet the scale of storage in Europe remains trivial compared with the expanding size of wind and solar power and in the UK companies are not exactly beating down the door to invest in new pumped storage projects.
Pumped storage hydro provides the most energy efficient, cheapest per MWh and most scalable storage option. How, for example, could hydrogen ever compete or be viable? With electricity – hydrogen – electricity round trip efficiency of the order 30%, at least 3 times as much off-peak electricity would have to be purchased than could be sold.
The massive Coire Glas pumped storage scheme in Scotland has been approved by government but still awaits an investment decision from Scottish and Southern Energy plc. Supposedly designed to store surplus wind power, I do not understand how they can make money storing water for several days awaiting a lull in the wind  but the substantially lower cost of this scheme (see below) may provide an explanation.
I’m uncertain if the paucity of new pumped storage projects is down to a lack of suitable sites or is it down to high capital costs? FLES – O-PAC may help resolve the former issue but it comes at a high price. Coire Glas comes with a price tag of €1.12 billion. With storage capacity of 30GWh (€37 million per GWh storage capacity) it is over three times larger than O-PAC (€214 million per GWh storage capacity). The cost of Coire Glas has been kept down by using a natural lake as the lower reservoir, but there is only so much pumping that can be done before the hydrology and eco-systems are wrecked.
In line with the thinking of Scottish Renewables, it seems that further substantial market incentives may be required to promote and expand renewable electricity storage .
 Flat-land Large-scale Electricity Storage; Global Energy Village 2015; Dr J. Huynen of SOGECOM.
 Energy Matters, Euan Mearns The Coire Glas pumped storage scheme – a massive but puny beast
 Energy Matters, Euan Mearns The Loch Ness Monster of Energy Storage
 Energy Matters, Roger Andrews El Hierro – another model for a sustainable energy future
 Energy Matters, Roger Andrews How Much Battery Storage Does a Solar PV System Need?
 Energy Matters, Roger Andrews A Potential Solution to the Problem of Storing Solar Energy – Don’t Store It
 Scottish Renewables Pumped Storage – Position Paper
 Valuation of Electricity Storage. Irina Zinkevich and Ruut Schalij, Energy Risk Advisory Group.