Energy Storage - a distraction for renewable energy? By David Milborrow
"Renewables need storage", claim their detractors -- and sometimes their supporters as well. "Storage can transform the economics of the intermittent renewables", proclaimed an august body recently.
Getting to grips with the issues
So why all the fuss? The idea, many argue, is that storage can be used to "level the output" of the variable sources of renewable energy. Whether or not that output really needs to be levelled brings us to the central issue. Consumer demands and power station outputs are actually more of a liability to the System Operators of large power systems than the relatively benign meanderings in the output from wind or photovoltaic systems.
Power stations "trip", to use the jargon, when their entire output is instantaneously lost. This may be due to instrumentation faults, short-circuits, steam turbines malfunctioning, or a variety of other causes. These trips are unpredictable and, as a result, power systems are always operated so that there is an adequate reserve of plant standing by, ready to take up the lost generation.
Similarly, electricity consumers can be fairly capricious and, although their habits are fairly well-known, can impose demand surges on electricity networks, when, for example, popular sporting events finish earlier than expected and half the nation rises to make a cup of tea.
Power systems: squaring up to uncertainty
System Operators, with the responsibility for maintaining electricity supplies, look at the overall balance between generation and demand and the associated uncertainties.
The level of uncertainty dictates the level of reserves, although various other factors come into the equation. To put things into perspective, when the demand on the network in England and Wales is 35,000 MW (round about the average level) then about 750 MW of Primary Response plant (the exact details need not concern us) will be required if the 1320 MW Sizewell B nuclear power station is operational.
Introducing a little bit of wind, or PV, on to the network in England and Wales will not upset these sums at all. What matters -- and this is the crucial issue -- is the impact that the variable renewable sources have on the overall uncertainty. Fortunately, there are a lot of data around, particularly from western Denmark, which sources around 20% of its electricity from wind, that enable the impacts to be quantified. With 20% wind energy on the network in the England and Wales additional reserves would indeed be needed, although they would not be "dedicated" exclusively to wind. If the extra costs of these reserves are charged to wind energy, then the premium would be a little under £3/MWh. However, that is for 20% wind; with 5% wind energy, the premium is just over £1/MWh, or just over 2% extra on the generation cost.
Electricity networks, wind and storage
Back to storage. Numerous utilities have looked at the issues surrounding the assimilation of the variable renewable sources and the unanimous conclusion is that it is the overall impact on the system as a whole that matters, not the particular characteristics of any particular plant. That will minimise costs to the consumers. Storage may well be beneficial -- perhaps even cost-effective -- as part of the strategy for operating the complete system as a whole, but isolating particular generating plant simply does not make economic sense.
Another way looking at the variability issue is to examine the way in which consumer demand and wind generation behave. Analysis of data from western Denmark suggests that, for 52% of the time, wind and demand are either rising together, or falling together. There is little point in sending wind energy to a store if it is helping the System Operator to cope with rising demand. Conversely, there is little point in drawing down from a store when wind output is falling, if the overall demand is falling.
So what are the prospects for storage? They are actually quite bright and have improved in the last few months due to the massive increases in the price of gas. The most fruitful way of using storage, however, is in the context of helping to manage integrated electricity systems. Storage can do two things: it can absorb energy when demand is low; that means prices will be low. That energy can then be fed back into the system at or near times of peak demand -- when the system will pay high prices for the energy. The other option for storage devices is for the operators to offer them as "reserves" to System Operators. This means that they can deliver power when there is an unscheduled trip at a power station or when domestic consumers are making too many unscheduled cups of tea. "Pumped Storage" systems -- where water is pumped to a high level reservoir and then channelled back when needed -- can respond very quickly and so demand premium prices.
Of the various storage technologies, pumped storage is the most firmly established, with around 22,000 MW installed in the United States and over 2000 MW in the UK. It is proven technology -- the turbines used are similar to those found in Hydro Electric plant, but they are reversible and pump the water uphill when necessary. The other advantage is that they can deliver power for minutes or even hours at a time; most other technologies struggle to deliver (or absorb) power for long periods.
In terms of installed capacity, compressed air storage follows next -- but some way behind -- pumped storage, with US capacity around 100 MW. As the title implies, compressed air is pumped into caverns and then released back -- usually into gas turbines. Research is in progress in Germany and elsewhere, apart from the United States.
Other storage technologies include batteries, flywheels and superconducting magnets. The latter have relatively low energy storage capabilities and high costs but "flow batteries" -- where chemical reactions in electrolytes take the place of the water in a pumped storage system -- show some promise. The "Regenesys" technology, developed by National Power, falls into this category. Although its new owner has discontinued the work -- perhaps confirming that it is difficult to make ends meet -- other, similar, systems are being developed in Japan and elsewhere.
It's all down to economics
Several utility studies have suggested that the target cost for storage systems is around £500/kW. It is difficult to be precise, as it depends what revenues can be realised from various "reserve" services. That assumes a pretty intensive use for the system, remembering that the maximum load factor of a storage system cannot exceed about 35-40% -- it spends half its time charging, and the overall efficiency will be around 80%, at the very most. Conventional batteries can meet this target -- but their storage potential is limited. Only pumped storage systems in very favourable locations can meet the target. Compressed air storage might be able to meet the target -- but exploitation is so far rather limited. A storage systems used solely to "level the load" of wind or solar power plant would almost certainly have an even lower load factor than 35% -- so it's energy delivery cost would be high. Given that the difference in value between "firm" and "non-firm"power rarely exceeds about £5/MWh -- about 1-tenth of that delivery cost, this shows why storage for variable renewables is unlikely to make economic sense. It does make sense for island situations, however, where the economics are completely different and the amount of worldwide research in progress suggests that further exploitation in large power systems will continue.
David Milborrow is an independent consultant who specialises in technical and economic issues relating to the renewable electricity sources, particularly integration issues. He has carried out assignments for clients in Britain, Europe and North America. He is on the board of the British Wind Energy Association and is technical adviser to the Journal "Windpower Monthly". An article discussing storage issues in more detail can be found in the "Wind Insight" section of the WPM web site: http://www.windpower-monthly.com/
David can be contacted at David.Milborrow@btinternet.com