I came across one or two proposals over the last few weeks for mega sized pumped storage facilities in Scotland to back up intermittent renewables (such as wind), which I thought it would be worth commenting on.
The first proposal involves taking Loch Morar, a freshwater loch 300m deep, but one just a few hundred metres from the shores of Loch Nevis (salt water). The plan would be to build a dam 280m high across a narrow pass to the North, which would allow an upper reservoir to be created 300m high. Water could then be pumped in either direction to create 600m of head and provide up to 1800 GWh’s of energy storage. Further options to build another dam to the south and pump water in and out of Loch Nevis would also exist.
By comparison, the UK’s current storage capacity is about 30 GWh’s. The UK’s dams produce a further 5,700 GWh’s annually, so if we assume say 45 days of storage on average that implies a further 700 GWh’s of unreplenishible storage.
And if this proposal seems a little oversized, there’s what is appropriately billed as “the loch Ness monster of energy storage”. This proposal is to build a dam 300m high to impound Strath Dearn, a glen in the Monadhliath mountains in the North Highlands, about 400m above sea level. Water would then be pumped along a canal from the Firth of Forth (given the obvious complications with pumping so much seawater such a long distance, I’d argue it would be easier to pump water up from nearby Loch Ness instead).
Either way the resulting reservoir would be capable of holding 4.4 billion m3 of water at a height of as much as 650m above sea level, representing a pumped storage capacity of 6,800 GWh’s. Enough to not only back up the whole of the UK’s renewables, but most of Europe’s!
Certainly there are some obvious criticisms of these projects. For example, at peak output the Strath Dearn facility would be capable of delivering +255 GW’s to the grid (the UK currently averages about 30 GW’s with a peak of perhaps 50 GW’s in winter), which raises the small issue of how to transport that much power South or North (when sending it for storage). You’d be covering most of the Grampian’s in powerlines!
There’s also the matter of pumping large amounts of sea water inland and the environmental effects of any leaks and spills. I would note that Sea water pumped storage is not a new idea, the Japanese have just such a plant operating on Okinawa, one they hoped eventually to become a string of such facilities to help their nuclear industry expand (nuclear reactors have essentially the oppose problem of renewables, they want to be on all the time, but grid demand fluctuates, meaning they need something to provide peaking power or some form of energy storage). While only 30 MW’s this facility has successfully operated for several years and many of the technical issues relating to leaks of sea water have been addressed.
I would also chuck in the need for another reservoir at the base of the dam, as a sudden swap from pump to power out (or visa versa) could cause problems, hence why relying on water pumped in from Loch Ness rather than sea water would probably be a better idea.
Certainly these proposals do get around one of the major myths you’ll hear put out about renewables – that we can’t store the energy and thus renewables can’t be relied upon. Of course this was always a myth put forward by those who don’t understand how renewables worked, nor indeed that all energy sources need some level of “backup”, nor that a number of renewable sources (Tidal, biomas, hydro, geothermal, solar CSP) aren’t intermittent and others such as solar and wind power tend to be complementary (i.e. if its not windy, its usually sunny, if its cloudy, its usually blowing a gale). In short, such issues are simply not going to be a problem for a future low carbon grid, assuming it has a good mix of different renewables, suitably spread out across the continent with good interconnection and use of “smart grid” technology.
Indeed my main criticism is just that these proposals are too darn big, we’ll simply never need that much storage capacity. And it won’t seem sensible to me to put all our energy eggs in one or two baskets. I suspect a series of storage facilities spread out across the UK and the rest of Europe would be a better idea. Indeed there are a number of proposals floating around for new energy storage facilities across Europe (although none on this scale admittedly).
For example a proposed 6.8 GWh sea water pumped storage facility on Glinsk mountain in Mayo, Ireland. There’s also proposals to identify a number of coastal valleys around the Irish coast and flood them in a similar manner to the above Loch Morar scheme, gaining 100-200 GWh’s of storage at a time. And this report from the European commission indicates how the EU’s PHS capacity could be increased significantly. In Turkey alone (hardly a country known for its heavy rain!) 3,800 GWh’s could be added.
While it is often said that there are no suitable sites for new hydroelectric plants in the highlands, this is not true of pumped storage or micro-hydro. Take a look at any map of Scotland and you’ll quickly identify several obvious sites in a matter of minutes. And some of these sites are being seriously investigated. SSE is for example pursuing a planned facility in Coire Glas (just over the hill from an existing pumped storage facility at Ben Cruachan), with consent for the facility granted in 2013. In Wales a proposed facility Glyn Rhonwy is also being taken forward.
So my suspicion is that lots of little facilities tucked away here and there, along with the conversion of existing hydro dams to pumped storage (such as Ben Lawers and Loch Sloy as both have quite a large head of water), would probably be a better idea than one mega storage facility. This study by Strathclydes ESRU suggests that 514 GWh’s could be added by converting existing high head reservoirs to pumped storage.
But how much pumped storage would be needed? Well keep in mind that electricity is only a small fraction of overall energy consumption (about 20%). A future energy grid will be devoting quite a sizeable proportion of its output towards things like heating (36%) and transport fuels (at least 30%). The heat load in particular will likely be obtained from a combination of boilers and CHP plants of various sizes, initially running on a combination of biofuels and natural gas (with presumably CCS) and later on hydrogen. This means that there will be a sizeable capacity of generating plant still available to back up wind or solar, at least over the winter months.
Storage of hydrogen underground is not a new idea, a facility in the US has stored large quantities of it underground in a salt cavern since the 80’s without any particular difficulties. Technical reviews by Bossell (2006, Does a Hydrogen Economy Make Sense?), Eric Wolf (2015, Large-Scale Hydrogen Energy Storage) and Camparani et al (2009, Journal of Power Sources Vol. 186, focuses mainly on cars, but it considers the full life cycle of H2 production, transport and storage) suggest hydrogen could be sourced via renewables, stored and transported to end users with an overall efficiency of 55-70%. Okay, not as good as pumped storage (64-74% once we account for an average UK grid efficiency of 92%), but when you consider that H2 would directly integrate with much of the existing gas network, it makes it a better choice of storage for winter heating fuel. Either way, this cuts down the proportion of winter backup needed by a considerable margin.
Let’s suppose that with the upgrading of existing hydro-electric facilities and the addition of some extra capacity, including one or two new reservoirs, give us 1.5 times the level of that 514 GWh’s figure mentioned before, that’s 514 x 1.5 = 771 GWh’s, plus existing capacity of 30 GWh’s gives about 800 GWh’s without breaking too much sweat.
Tidal lagoons (such as the scheme proposed in Wales) can also provide energy storage, Mc Kay (not one of my favourite people, he has a habit of getting his sums wrong), proposes arrays of these around the UK coast which could provide 20 GWh’s of storage a go, along with 400-650 MW of installed capacity. So 10 such facilities would provide 200 GWh’s of storage.
Its safe to assume that more cars in the future will be electric, keeping in mind that even hybrid cars will have a battery with at least 10-20 kWh’s each of capacity and +40 kWh’s each for fully electric cars. This means that a modest fleet of say 6 million such vehicles (out of an existing UK fleet of 29 million cars), assuming we only use the first 10-20 kWh’s of the battery capacity, would be able to supply between 60 – 120 GWh’s of storage.
There are several new energy storage technologies in development. Large battery arrays to back up renewables is one option. Elon Musk, amongst others, is seriously investigating these as a viable energy option, although it is early days. Another is the idea of CAES (compressed air energy storage) or the recent innovation of LAES (liquid air energy storage). LAES is an interesting idea, because it can be easily added to existing thermal power infrastructure, something as noted we’ll have no shortage of, without too much additional cost. Its storage energy density of 100-200 kWh’s per m3 makes for pretty compact storage.
Of course these are all emerging technologies, some won’t work out and its hard to tell which will succeed (another days article). But let’s suppose we can get the same output from them as we can from the cars. So all together that gives us at least 1,000 – 1,200 GWh’s worth of storage across the UK, without trying very hard. We also have the capacity of existing hydroelectric reservoirs of approximately 200 GWh’s (once we account for those reservoirs converted to PHS), although remember this is a one shot source than will take weeks to replenish. So all in we’ve got up to 1,400 GWh’s of storage to play with.
Would this be enough? Another study by the ESRU suggests that even with a grid drawing on 60% of its power from variable renewables, 550 GWh’s would provide adequate backup for at least seven days. This would seem to imply that the level of storage I’m proposing would provide at least twice the amount actually needed. Although I would note that this study probably doesn’t account for GW’s needed (i.e. sudden draws on the grid at inconvenient times) or extended periods of low renewable output.
As a result some would say no, the UK’s average daily draw of electricity is about 900 GWh’s, so this would seem to imply the above would provide just a day and a half worth of storage (a bit more than half a day if we believe Strathclyde Uni!). However, its worth remembering that this implies an “all stop” scenario with no power being generated and everything turned on. i.e. No sunshine, no wind, no waves, no tides (which is going to be difficult without the moon disappearing!), all the thermal plants down, whatever is left of the UK’s nuclear fleet out to lunch (then again, they have been knocked offline by storms or heatwaves). And of course we’re looking at the UK in isolation. One assumes a future grid will be connected, to the continent, Ireland and possibly even to Norway (and its dams). So we’re imagining a scenario where all of these sources are also unavailable. And we’re ignoring existing policies which calls for “load shedding”, whereby major industrial users of power de-rate or go offline to help ease the pressure on the grid at certain times. Also there are other means of energy storage in the UK we’re not accounting for above. For example, a report by the ERP mentions the nation’s hot water tanks hold at least 65 GWh’s.
The idea that all of these sources could all go down at the same time as all of the loads are on is a pretty unlikely scenario and it’s highly improbable to last any more than 24 hrs. Keep in mind that if the existing grid faced the sort of scenario listed above, we’d be in trouble pretty quickly. Much of the UK’s electricity and heat energy is drawn from the gas reserves of the North Sea. An interruption to pipelines coming off the North sea, given that the UK has very little gas stored onshore, would leave the country with just a few days supply at winter withdrawal rates. If France developed some issue with its nuclear reactors (a safety scare of some sort) and was forced to shut many of them down, they would have problems straight away. As would parts of southern England who depend on French nuclear for electricity.
In short, we’re placing design requirements on a future grid that exceed those of our existing grid. And indeed one could argue the existing UK grid is itself overdesigned. As someone from the national grid once put it to me, how much it would cost to adequately back up the grid for intermittent renewables to become the majority source of electricity is like asking how long is a piece of string. It boils down to the question, how reliable do you want the grid to be? 100% reliable all of the time? or 95% reliable most of the time, but for half the price?
Keep in mind that in other parts of the world brown outs or blackouts are far more common. In the US for example, while its not often the power goes off, it does happen from time to time (heatwaves, forest fires, ice storms, tornadoes, price gouging utilities, etc.). Sufficiently often that anyone who needs power 24/7 generally has a standby generator on-site. Now the Americans could easily add enough spare capacity into their grid, put the cables underground, etc. and have a grid that works all of the time. But would people be willing to pay for that? Probably not.
Consequently about the only situation where I could see such mega storage systems being built is part of some sort of strategic energy reserve (at which point any sensible analysis based on economics and actual requirements goes out the window as we’re talking in terms of national security). Or if there was a failure to deliver on hydrogen as a substitute for natural gas and instead rely on electricity. This would then require a large reserve of electricity to meet winter heating needs. In this scenario the withdrawal rates would of course be much lower, probably in the order of ten’s of GW’s (i.e. capacity of the Three Gorges Dam), well within the limits of a few HVDC lines to support.
Either way, what these proposals do show is that necessity is likely to be the mother of invention. Even if we did need the sort of vast storage levels the naysayers suggest (and we won’t) it would still be technically possible to provide such levels of storage with existing technology, nevermind the potential storage capacity provided by future advances in technology.
That said, any level of storage, regardless of the technology used, is going to take time to develop plan and build. The major problem with that is, if you listen to the IPCC, time isn’t exactly an asset we’re in an abundant supply of.