This will to have to change. Without cheap, robust and very large scale electricity storage, electricity grids are going to find it very difficult to cope with the unpredictability of vastly greater supplies of electricity from wind, wave or sun. HIghview Power, a UK company that has operated in what private equity calls ‘stealth mode’ for several years, went public yesterday with an intriguing proposal for a new form of energy storage – air liquefaction. The energy commentators read the press release and politely yawned. Were they right?

The economics of this technology look interesting. What is even more compelling is that you could bolt together a large plant using conventional components freely available today from a variety of major suppliers. Unlike some of the really wacky suggestions for storing energy, we pretty much know that Highview’s ideas will work.  A 350 kW pilot plant alongside the Slough power station has been through extensive testing for the last six months or so.

So how does it operate? You take ambient air and put it through a liquefaction plant using electricity. (Hundreds of these plants around the world today make liquid nitrogen, oxygen or natural gas).  Liquefaction works by expanding a gas, which causes its pressure, and thus its temperature to fall. This technology is a hundred years old. The process uses substantial amounts of energy.

Allowing liquid air to expand increases its volume many hundred fold. This will produce high pressure in any sealed container. If the gaseous air is allowed to escape through a turbine, electricity can be generated. This second phase produces about 55% of the input energy, says Highview. This relatively low number can be improved to perhaps 70% by using waste heat from nearby  industrial processes, such as the hot water from the cooling processes in a nuclear or fossil fuel power station.

How does efficiency this compare? Here are some very rough figures for other means of storing electricity.

(A previous article on Carbon Commentary assessed the economics of using stored hydrogen for electricity production).

The huge advantage of Highview’s plant, if it works as planned, is that each of the main alternative storage technologies have intrinsic problems. Hydrogen is inefficient and the equipment is expensive. Compressed air requires large amounts of storage. Lithium batteries are expensive and don’t like being discharged too often. (Other battery systems are less problematic but they have other disadvantages). Pumping water uphill is cheap and well-understood. There just aren’t many places where it can be done economically.

Highview quoted me a figure of £1,000 per kilowatt of output power. Let’s be clear about what this means. The Slough pilot plant can produce 350 kW of electricity. So the cost of a commercial plant would be about £350,000. (The cost of the pilot was much greater, of course). The Slough kit can deliver about 2.5 megawatt hours when fully charged. That is, it can work for seven or eight hours at full power. If it can be achieved, £1,000 per kilowatt of electric power is highly competitive with most other storage technologies, particularly since operating costs are so low.  A large pumped hydro plant would be comparable, but hydrogen could be four or five times as expensive.

Build an air liquefaction plant and expansion plant to Highview’s designs and what do you get? A megawatt plant would have a capital cost of a million pounds or so. To be cautious, we’ll assume £2m. This plant can be used in several different ways.

First, it can respond to short-term grid problems. Highview says it might take the plant a couple of minutes to start producing useful power to respond to a power station failure or grid problem. This isn’t quite fast enough for the real (but rare) emergencies when gigawatts suddenly disappear from the grid. But is good enough to help respond, for example, when wind farms start having to close because of excess wind speeds. The reverse situation, when National Grid has to pay wind operators to close down because of an excess of national electricity supply, can be addressed by Highview plants. They can also absorb surplus power while the generator close down other sources of supply.  National Grid pays for the small producers, such as the emergency diesel generators at hospitals and sewage farms to be available to produce electricity at less than half hour’s notice. The disadvantage is that these facilities can’t take in surplus power whereas Highview’s plants can act like batteries, either taking in electricity or discharging.

Second, they can provide extremely useful ‘peak shaving’. Electricity demand varies throughout the day. Individual large customers pay both for their total usage of electricity and for the amount of capacity they are using at the times when the total UK demand is at its peak. Such a customer might invest in Highview’s system to reduce its annual capacity payments. Each time the real-time grid information indicated that a peak in electricity was being reached (usually at around 5pm in the winter), the air liquefaction plant would be switched to electricity production, minimising the peak demand of the user and hence reducing the payment for maximum capacity used. In countries such as South Africa, which have electricity grids that are sometimes unable to cope with peak demands, Highview technology could be particularly useful.

Third, the plants could use electricity when it is very cheap and sell it when it is expensive. Typically this means storing power at night and then discharging at the early evening peak. But remember that if the efficiency (input electricity versus output electricity) is only 55%, the price difference will have to be large to make this worthwhile.

What the plants cannot do – because they will never have enough capacity to work uninterruptedly for days – is to replace wind power at times when the turbines are stalled because of metrological conditions. We will always need to have gas turbines for these events. Nevertheless, air liquefaction looks to be potentially the cheapest and most robust way of adding the several gigawatts of energy storage capacity that the UK grid needs if it is to deal with the unpredictability of 10,000 offshore wind turbines. At present, I suspect the financial calculations won’t quite provide the incentive to make the investments necessary. But  the future electricity market will only work if there is a strong financial signal that encourages storage investments. It must happen eventually. Venture capital really should be interested.

Both of these assumptions are probably wrong. One credible source sees the cost of offshore wind falling to levels competitive with gas, albeit over several decades. And foreign interest in offshore wind is growing as the best onshore sites are completed. A Chinese study estimated the potential for exploitable wind power offshore is about 750 gigawatts, perhaps ten times the UK’s likely resource. Over the next few years China plans enormous investments in sea-based turbines. Similar opportunities are available in the US.

First, a couple of points as to why should the UK want to specialise in offshore wind. The country’s territorial waters are blessed with relatively high wind speeds compared to even the best onshore sites. The UK’s resources are about 40% of the total wind power available to Europe. Installation of turbines is difficult but the UK is well placed because of its expertise in putting oil and gas platforms safely in place in deep, rough water.

Early development of offshore farms has been expensive in the UK (and elsewhere) partly because of difficult construction conditions, a shortage of fixing vessels, limited competition between offshore turbine manufacturers and low levels of historical reliability of installed equipment. The push to develop larger and larger farms, usually with increasingly large individual turbines, should reduce capital and operating costs as operators get more experience. Perhaps as importantly, a large number of turbine manufacturers are in the process of introducing new models suitable for the rough UK conditions. The scope of steep reduction in costs is certainly present, a point denied by the Policy Exchange author Simon Less.

The engineering consultants Mott MacDonald provided an estimate for the recent Committee on Climate Change report on low carbon electricity. The consultancy gives the following prospective figures for 2040. Gas with carbon capture (CCS) is probably the least costly way of generating  electricity from fossil fuels without adding significantly to CO2 concentrations.

Offshore wind – £60-£96 per MWh
Gas with CCS – £95-£104 per MWh

(Numbers on pages 7-9 and 7-10 of http://hmccc.s3.amazonaws.com/Renewables%20Review/MML%20final%20report%20for%20CCC%209%20may%202011.pdf

But you might well ask whether any technology that takes thirty years to get to cost competitiveness is worth backing to that point. The answer is that a large number of expensive wind farms will have been put in place before the experience gained reduces the cost to reasonable levels. (For information, the current wholesale price of winter electricity is about £60 per megawatt hour). Yes, it might take about £100bn to get to the Mott Macdonald 2040 figure but the issue we face is that no technology –other than onshore wind – is likely to be much better. And getting tens of thousands of onshore turbines across all the UK’s western coasts is not looking politically feasible.

The Policy Exchange recommendation seems to be that we should spent a lot more on basic research in low carbon technologies.  However the arguments why this would achieve faster and cheaper results than a hard-nosed push for cheaper offshore wind through heavy subsidy of early turbine parks are simply not made in the think tank’s paper.

The more obvious error is to assume that no other countries are particularly interested in offshore wind.  Having opened its first intertidal wind farm just three weeks ago, China says it wants 30 gigawatts of offshore wind by 2020. The exploitable resources of 750 gigawatts compares to the 30-35 that the UK has plans to develop in the next decade or so.  That’s right, China alone sees a market twenty times the size of the UK.

Recent semi-official suggestions are that the cost of the Chinese intertidal farm will run at about £80 a MWh are probably highly optimistic, but show what might be achieved elsewhere in shallow waters.

The US has a similar sense of the value of offshore wind resources, with a figure of just over 1,000 gigawatts being seen as possible at sites with wind speed of more than 7 metres a second average wind speed and in water less than 30 metres deep. Total potential resource might be four times as much, approximately enough to power the whole US at capacity factors of 30%. (It should be admitted that progress in actually building the wind farms off the US coast has been lamentably slow and dogged by controversy. An excellent site off Cape Cod has been blocked by powerful local residents for years).

In Europe, the UK leads in offshore wind but other countries continue to invest in new turbines. 235 wind turbines were installed in European waters in 2011, averaging over 3 megawatts each. Germany, Sweden, Belgium, Denmark, the Netherlands, and Finland all now have offshore wind farms. The German decision to abandon nuclear virtually obliges it to focus on Baltic wind farms as the most significant source of low carbon electricity over the next ten years. The European Environment Agency says total EU installed offshore wind will rise 17 fold by 2020.

The list of other countries beginning to develop wind grows by the month. S Korea has just announced its first major play, a 2.5 gigawatt farm off the south-western coast. Canada has recently announced firm plans for a pathbreaking development off Ontario.

The UK’s enviably rich offshore resources and its leading world position in the development of complex wind projects miles from a coast give the country a major set of potential advantages in exporting construction and engineering skills around the world. The relentless negativity about wind from an intelligent think-tank is disappointing.

This assertion is entirely based on the work of Colin Gibson, a former National Grid engineer, who has made some informal estimates of the cost of integrating wind power into the electricity networks. He suggests that these costs are about £60 a megawatt hour, adding perhaps 70% to the cost of electricity from wind turbines. Ms Lea fails to mention that many, many other analysts and engineers have also estimated the extra costs of adding large volumes of wind power to the electricity system. In this note I suggest that these alternative sources support a view that Mr Gibson’s estimates are wrong by about a factor of four, meaning that Ms Lea’s contention that wind is a very expensive technology is based on shaky foundations.

The task of estimating the relative costs of electricity generating technologies is complex. The result depends critically on the assumptions we make about the cost of investment capital, the amount of bank debt that can be used, how long the generating plant takes to build, the cost of fossil fuels and a host of many other variables. The final numbers, usually expressed as pounds per megawatt hour of electricity produced are, at best, approximations.

Ms Lea uses as her source the figures produced by Mott McDonald, an engineering firm, in 2010. She should probably have the used the more tentative and up-to-date figures generated by Mott McDonald for the Committee on Climate Change in 2011. The 2011 numbers give ranges of estimates for the direct costs of all the main technologies, for both today and in the future. These figures suggest that onshore wind power is broadly competitive with nuclear power. Offshore wind is currently much more expensive but advances in technology are projected to make it competitive over the next few decades. Mott McDonald, whether in 2010 or in 2011, certainly doesn’t see direct costs of wind power as ‘quite uneconomic’ and, to be fair, neither does Ruth Lea.

Wind power is more costly to integrate into the grid than conventional power stations. There are three major types of extra charges and these incremental costs are not included in the Mott McDonald figures.

• The impact of having to have spare capacity on hand to react to unexpected changes in the outputs of UK wind farms. (Even if the electricity network were entirely powered by large nuclear plants, the UK would still need this spare capacity, ready to ramp up to full power, because of the risk of a station ‘tripping’ and its power not being available to the National Grid. Wind farms are actually less risky than a single nuclear power plant)
• The cost of having to construct power stations that are used only when the wind is not blowing.
• Charges arising from having to construct new distribution lines to connect wind farms, often in remote locations or offshore, to the National Grid.

Mr Gibson’s work, on which Ruth Lea entirely relies, suggests that the cost of these extra measures is about £60 per megawatt hour.

Table 1

Other sources give very different figures for the unseen costs of wind generated electricity. From the many available, I have used two reports produced by consulting engineers and by electricity network specialists. As far as I can see the numbers in these reports are representative of the consensus view of wind integration costs.

I don’t claim that these numbers are right, but I do think that Ms Lea should have given reasons why this recent work is less appropriate to use than the rough estimates of a single individual, however competent.

Table 2

(1)    Sinclair Knight Merz, Growth Scenarios for UK Renewables Generation and Implications for Future Developments and Operation of Electricity Networks, June 2008. (A report for BERR, now the Department of Business, Innovation and Skills.) Page 90

(2)    Sinclair Knight Merz, Growth Scenarios for UK Renewables Generation and Implications for Future Developments and Operation of Electricity Networks, June 2008. (A report for BERR, now the Department of Business, Innovation and Skills.) Page 91

(3)    Energy Networks Strategy Group, Our Electricity Transmission Network: A Vision for 2020, March 2009. This report estimates the gross cost of new transmission infrastructure to cope with dramatically increased renewables generation at £4.7bn. I turned this into a cost per megawatt hour using the calculator in Mr Gibson’s spreadsheet, thus ensuring reasonable comparability with the figure used in Ruth Lea’s paper.

The implication of the far lower costs in Table 2 is that we should add about 15-20% to the direct costs of wind power to properly account for the impacts of this source of electricity on the costs of the network as a whole, not 70%. This leaves wind as an entirely economic and carbon-saving technology. Did Ms Lea, a noted climate change sceptic,  use Colin Gibson’s very high figures because of her dislike for the renewables policy of the UK government?