And tempest roars, with tempest vying,
From sea to land, from land to sea,
In their alternate furies tying
A chain of deepest potency
There are many things I dislike about renewable energy, or – to be more precise – about the industry that promotes it. I dislike the misleading claims its advocates make. I dislike the tokenism that attends it: a petrol company might put a wind turbine beside a filling station (because it is too far from the grid to make a connection worthwhile) and its customers think it has gone green. I dislike the way in which covering the countryside with wind turbines is often seen as a solution to our excessive consumption of fossil fuel, as if the new technology, in the absence of policy, replaced rather than simply augmented the old one. I dislike the way in which environmentalists sometimes choose to overlook the destructive effects of renewable technologies, such as the tidal barrages which drown the ecosystems of estuaries.
But even in my grumpiest moods, I can also see its virtues. The wind, waves and sun are not going to run out – or not while we still occupy the planet. Neither Mr Putin nor any other energy monopolist can switch them off. No wind farm can ever melt down, or present a useful target for terrorists (modern Don Quixotes notwithstanding). Decommissioning is cheap and safe. The energy required to build the machines on the market today is a small fraction of the energy they will produce,2,3 and as soon as that has been accounted for, they emit no carbon. While renewable technologies can dominate a landscape, this impact is surely less significant than the destruction of the biosphere. But even if we were to overlook its damaging or counterproductive effects, there remain good reasons for questioning the claim that our electricity could be supplied wholly or even largely by renewable power. As I hope you are discovering, nothing in this field is simple.
The first question that people who advocate the replacement of our power stations with renewable energy must answer is this: is there enough of it? The United Kingdom – islands surrounded by high winds and rough seas – has the best resources in Europe. But even here there is room for doubt about whether ambient power would be sufficient to meet our demands.
We use around 400 terawatt hours of electricity a year.4 A terawatt hour (TWh) is energy produced at the rate of 1 trillion joules per second for one hour. That may not make it much more comprehensible, but you need bear only the comparative figures in mind. In 1999, a consultancy called the Energy Technology Support Unit was commissioned by the British government to work out how much renewable energy the country has. The table overleaf gives the figures it produced for what it calls the ‘practicable resource’. Practicable resource means the power that could be produced for a reasonable cost after various constraints (such as not building in national parks or where the seabed is too soft, or where turbines would interfere with radar or migrating birds) have been taken into account. A ‘reasonable cost’ was 7 pence per kilowatt hour (kWh) or less. Keeping costs down to this level means that the machines can be built only where the resource is concentrated – in other words where the wind blows hard or the tide flows strongly – and only where they can be clustered. Scattered wind turbines, for example, would cost a fortune to connect to the grid.
You might notice that some technologies, such as tidal barrages (a dam across an estuary), are not listed in the table. This is because the consultancy concluded that the practicable resource is zero – something I am rather glad about. So, astonishingly, it looks as if even in the windiest, most battered nation in the European Union,
energy source |
practicable resource in 2025 |
onshore wind |
8 |
offshore wind |
100 |
waves |
53 |
tidal streami |
2 |
solar photovoltaic |
0.5ii |
hydroelectric |
7iii |
Total |
170.5 |
i. ‘Tidal stream’ means rotors making use of free-flowing current, as opposed to ‘tidal barrage’, which means a dam across an estuary. |
|
ii. ETSU’s report is extremely confusing on this point. It starts off by claiming that ‘the maximum practicable resource in 2025 is 266TWh/year [TWh means terawatt hours]. The maximum practicable resource includes PV applied to all orientations.’ This is – to say the least – an eccentric proposition: that solar panels can be fitted to building surfaces at all points of the compass. It then goes on to claim a ‘technical potential’ in 2025 of 37 TWh, and a ‘market potential’ of 0.17. Then it produces a ‘resource-cost curve’ showing how much electricity can be produced at 7p per kWh or less, which gives 0.5 TWh at an 8 per cent discount rate. This appears to be its final figure. The Royal Commission on Environmental Pollution seems to interpret its report as expressing pessimism of this order, when it says that ‘ETSU estimates that there will be no photovoltaic resource available at a cost of less than 7 p/kWh by 2010, and only a very limited amount (average output of 0.2 GW) by 2025.’ But I cannot find the 0.2GW figure (which is actually a measure of capacity, not output) in ETSU’s report. |
|
iii. 4TWh is already in operation. ETSU suggests that the only significant remaining resource whose exploitation would not cause serious problems is 3TWh in Scotland. |
|
Source: Energy Technology Support Unit.5
we could harness less than half the ambient energy we’d need to produce our electricity.
It is now clear, however, that these estimates are far too pessimistic. The Energy Technology Support Unit (ETSU) assumed that the national grid remains unmodified – in truth it could be extended into new areas. The costs of some renewable technologies have already been falling much faster than it proposed. It investigated the potential for wave power in just five places: in reality we could use it over a much wider area.
In 2005 a government paper called Offshore Renewables – The Potential Resource produced a number for what it called the ‘Potential offshore wind generation resource in proposed strategic regions’. This it estimated at 3,213 terawatt hours: over eight times our total electricity demand.6 What makes it even more astonishing is that the ‘proposed strategic regions’ are all off the coast of England: the figure does not take account of the seas off Scotland, Northern Ireland and Wales, where the wind is generally stronger (but, because the water is mostly deeper and rougher around these countries than around England, it is harder to plant turbines there).
The paper does not explain how much of this ‘potential resource’ could become a ‘practicable resource’. But a likely estimate is between 5 and 10 per cent. So if we could cut our demand by 25 per cent – reducing the 400 terawatt hours we use each year to 300 – offshore wind power alone could, in theory, supply somewhere between one half and all of our electricity.
One of the reasons why the government’s new estimates are so much greater is that offshore wind turbines are already bigger than ETSU envisaged. ETSU suggested that each turbine would have a capacity of 1.5 megawatts, while the government’s paper proposes 3 megawatts. In fact turbines of this size are already being installed. Some people are now suggesting that by 2030 10-megawatt machines could be built.7 For the same reason, we can regard ETSU’s figure for the onshore wind resource as too low, as it assumed that the machines erected on land would have a capacity of only 600 kilowatts.8 Already 2-megawatt windmills are being built on some of our hills. This seems to be the maximum size for onshore turbines however, as there are no lorries big enough to carry longer blades.
ETSU also assumed that no wind turbine would be built in water deeper than 40 metres. The government believes that they could be planted in the bed of seas as deep as 50 metres. It also envisages that they could be built much further from the land. If you built a very large wind farm 100 kilometres, rather than 20 kilometres, from the shore, its paper shows, the extra expense of attaching it to the national grid would add only about 5 per cent to the total cost of the project: ‘It is therefore likely that the industry could consider potential sites well outside territorial waters, and perhaps as far as 100 kilometres offshore.’9
Though it does not mention them by name, the government seems to have discovered something that has also captured my attention: High Voltage Direct Current cables. I am coming to believe that they could change the world.
The first electricity networks, built by Thomas Edison, carried direct current (DC) electrical power, but only at low voltages. This meant that they were restricted to short distances: a maximum of about two kilometres. Alternating current (AC) could be generated and distributed at high voltages, and soon became the means by which electricity was transmitted almost everywhere. The pylon lines upon which we all depend and which we all detest are – almost without exception – high voltage AC. But the development of the high-voltage valve and much lighter wires, which permit the long-distance transmission of direct current, changes the formula.
Setting up a high voltage DC line is, in the first instance, more expensive than setting up a high voltage AC line. The initial loss of electricity when you hoist it onto the transmission system is greater.10 But the longer the transmission distance, the better DC begins to look. DC pylons are smaller, so the transmission network requires less land (and is not nearly as ugly). A graph published by the World Bank shows AC costs exceeding DC costs beyond about 650 kilometres.11 This distance is falling rapidly, because the costs of DC systems are plunging, while the costs of AC systems (due to their higher environmental impact and requirement for land) are rising. The World Bank claims that new DC cables made of extruded polyethylene could now make economic sense over as little as 60 kilometres.12 But most importantly, though the initial electricity loss on a DC line is higher, it does not increase with distance. On AC systems, by contrast, the longer the line, the more you lose. There is no inherent limit on the length of a DC cable. Already there is a line in the Democratic Republic of Congo that is 1,700 kilometres long.13
What this means is that you can draw your electricity from a far greater area than before. High voltage DC, which can be run along the sea bed, opens up any patch of sea shallower than 50 metres to wind turbines, and pretty well all the continental shelf to wave power devices, which (because they float) can be anchored at greater depths. Since wind speeds rise by around one metre per second with every 100 kilometres from the shore,14 this means that the cost of renewable power could actually fall with distance from the coast.
While offshore wind power is currently more expensive than onshore wind, the economies of scale permitted by massive developments attached to long-distance cables means its price could fall very rapidly: according to one estimate, by 40 per cent in ten years.15 And, beyond a certain distance, the only people whose aesthetic sensibilities are likely to be offended are trawlermen. You can install wind turbines which rotate faster (and are therefore both noisier and more efficient) without upsetting anyone.16
But it’s not just new wind and wave power that the longlines could exploit. At the moment there is an inverse relationship between the availability of solar power and human habitation. It is most concentrated and most reliable in deserts. For years, rogue environmentalists have been pointing out that solar electricity generated in the Sahara could supply all of Europe, the Gobi could power China, and the Chihuahuan, Sonoran, Atacama and Great Victoria deserts could electrify their entire continents. These people have been dismissed as nutters. The development of cheap DC cables suggests that they might one day be proved right.
There are several technologies with the potential to produce cheap electricity in very sunny places. One involves the mass production of solar photovoltaic cells of the kind now being installed, in small panels, on people’s roofs. On a small scale, at high latitudes, solar electricity is – and will probably remain – more expensive than almost any other kind of electricity. But huge photovoltaic farms in the desert have the potential to realize economies of scale which have hitherto been unavailable. Perhaps even more promising is a technology called ‘solar thermal electricity’.
You can use a reflective dish or trough to focus heat from the sun onto a tube containing water or another liquid. This reaches a temperature of about 400°, and the steam drives an engine which generates electricity.17 Power farms using this technology have been operating in southern California since the early 1980s. Their output costs between 12 and 15 US cents (about 7–9 pence) per kilowatt hour.18 At times of maximum demand in southern California, electricity now costs between 10 and 18 cents.19 (Because of air conditioning, demand is greatest there when the sun is hottest, so the maximum output of solar electricity happens to match the maximum demand.) There are now similar schemes in Spain, Italy, Morocco, India, Mexico and several other countries.
Or you can use mirrors to bounce the sunlight onto a central receiver. The mirrors track the sun and focus its heat onto a small ceramic plate on a tower about 100 metres high. The plate reaches a temperature of 1000°, again generating steam which drives a turbine.20 A demonstration plant is already working in Spain. With temperatures like this, it should be possible to store some of the energy the plant produces in the form of molten salt, which would allow it to generate electricity even when the sun is not shining.21 Another scheme, which has also been tried at an experimental level in Spain, uses heat from the sun to drive air through turbines in a hollow tower. The Australian government has shown great enthusiasm for a solar ‘power tower’ 1 kilometre high in the outback north of Melbourne,22 but I suspect its interest is prompted more by the prospect of building the world’s tallest structure than by the project’s commercial viability, which is currently not demonstrable.
The International Energy Agency calculates that if solar photovoltaic panels were used to cover 50 per cent of the land surface of the world’s major deserts, they would produce eighteen times as much energy – which means 216 times as much electricity* – as the world now uses.25 In other words, you would need to cover only 0.23 per cent of the land to meet demand, assuming that people used electricity only when the sun was shining. In 2003 the agency suggested that this electricity would cost between 9 and 11 US cents per kilowatt hour, and that this would fall by between a quarter and a half by 2010.26
What makes the use of long-distance cables particularly exciting is that they allow you to obtain electricity from several different regions at the same time. This makes renewable energy much more reliable than it would otherwise have been. There might be a flat calm in the North Sea, but a steady breeze 100 miles off the coast of Ireland. When the sun in the eastern Sahara is going down, the sun in the western Sahara is at full strength. Long-distance cables, as well as opening up new sources of electricity, might also help to overcome the greatest constraint limiting the use of ambient power – its intermittency.
I have mentioned the reductions in cost which could be achieved as a result of economies of scale. But in another respect renewable power behaves very strangely: the more there is, the more expensive one of its components becomes.
The problem comes down to this: that the wind does not blow and the waves do not rise all the time. Unlike power stations which burn fossil fuel, they cannot be turned on when we want them. If we switched our entire electricity generating network over to variable sources of renewable power, without building a massive energy storage system, then whenever the wind or waves dropped, the grid would collapse. What this means is that our renewable energy needs to be supported by other forms of power.
All electricity generating systems have spare capacity. When a steam turbine trips out or a transformer fails, or there’s a sudden and unexpected surge in demand, the extra plant, which was either lying idle or operating at half-throttle, can be fired up to fill the gap. In the United Kingdom, for example, where the maximum demand for electricity requires 62.7GW of generating capacity, we have 75.5GW of plant.27 (Only some of this spare capacity can be used, however, as it includes generators that are out of service.)
The variable nature of wind and waves and other renewables means that though we might build many gigawatts of renewable generating capacity, we cannot retire a corresponding amount of power stations that burn fossil fuel. We will have to carry the cost of maintaining them and even, in the future – when the old plants die – building new ones, whose main purpose would be to sit and wait for the wind to fail. The more renewable installations we build, the greater is the impact on the grid of the wind ceasing to blow or the waves ceasing to rise. So each additional gigawatt of renewable power we install displaces a smaller proportion of conventional plant.
If we built 8 GW of wind farms, for example, they would allow us to shut down around 3 GW of power stations which burn coal or gas.28 But if we built 25 GW of wind (which would produce about 20 per cent of our electricity), we could shut down only 5 GW of old plant.29 Beyond that point, very few old power stations could be closed, however many windmills we built.30 Graham Sinden of Oxford’s Environmental Change Institute has shown that a more reliable mixture – of wind, waves and tidal power, rather than just wind – would allow us to raise the retirement rate a little: 26 GW of renewables could shut down about 6 GW of old plant.31 Even so, this means we would have to sustain 96 GW of power generation in order to provide the electricity we now obtain from 76.* It’s important to note, though, that while the capital costs of the electricity generating network would rise, the cost of buying fuel would fall, as the gas or coal burning plants providing back-up power for the new wind farms would be used much less often that they are today. As longas wind power pays its way, the fact that our total generating capacity is bigger doesn’t matter.
A report commissioned by the British government shows that if we obtained 20 per cent of our electricity from renewables in 2020, the extra ‘system costs’ would add between £140 million and £400 million to a total generating cost of £9 billion. If renewables supplied 30 per cent of our electricity, the system costs would amount to an extra £330–920 million.32 The great majority of these extra costs would be for ‘balancing and capacity’, which means keeping the old plants alive and firing them up when they’re needed. Even so, as a report by the UK Energy Research Centre suggests, if 20% of our power came from renewables, the extra system costs of a massive increase in the number of wind turbines are likely to be far lower than the possible economies of scale.33
And beyond 20 or 30 per cent? Well, hardly anyone even seems to be asking the question. I still haven’t got to the bottom of why this is. The Royal Commission on Environmental Pollution complains that ‘there appears to have been no research as yet’ into the question of how much renewable electricity the grid can take, ‘either by the National Grid or by any other body’.34
I suspect that researchers won’t go beyond 20 per cent or 30 per cent because no one is asking them to do so. The academic study of carbon reduction in the rich nations is blighted by government policy. Governments (which provide most of the money for research) set targets (in the UK’s case, 20 per cent renewables by 2020) and then commission people to find out how they could best be met. They do not commission them to find out whether they are the right targets, or whether they bear any relation to the technical and economic limits. So beyond 20 or 30 per cent, we are groping about in the dark.
Building more wind turbines and wave machines and other renewable generators will not allow us to shut down most of our existing power stations. But this does not mean that our remaining thermal plants will be burning as much fossil fuel as they do today. Those which are retained for the purposes of insurance will, for the most part, be fired up only when demand is high and the wind starts to drop.
The power stations which respond fastest or cost least to start either burn coal or burn gas in an ‘open cycle’ turbine.35 These are both inefficient means of generating electricity, so the carbon cost of wind is a little higher than it might otherwise have been. Moreover, more plant than before will have to be kept ‘part-loaded’. This means it is turning over at reduced capacity but ready to be taken up to full power almost instantly. Power stations running at part-load are 10–20 per cent less efficient than power stations running at full load.36
According to the Royal Commission on Environmental Pollution,
Although the total capacity of those [backup] plants would be substantial, it would be used only infrequently, and the resulting addition to the annual total of UK carbon dioxide emissions would not therefore be large.37
But because it does not consider any contribution from renewables beyond 20 per cent, we don’t know whether this would still be the case if we started producing, say, 50 per cent of our electricity by these means. A paper published in the journal Energy Policy goes into a little more detail:
Taking a conservative estimate of 10 per cent for the reduced efficiency [of part-loaded power stations]… the emission savings from the wind will be reduced by a little over 1 per cent. This can be compared with the 20 per cent of fossil fuel avoided by using wind generation.38
In other words, almost 99 per cent of the electricity produced by wind power would be carbon-free. But it also leaves us in the dark if we want to know what happens beyond 20 per cent.
Is this assessment reliable? Writing in the magazine Civil Engineering, the energy consultant Hugh Sharman points out that forecasters are currently unable to predict wind speeds on the following day to a higher degree of accuracy than 1.5 metres per second.39 Whether the wind is blowing at 7.5 metres per second or 9 makes a big difference to the output of a turbine: Sharman suggests that the difference corresponds to 21 per cent of its total power-generating capacity. The wind’s unpredictability he says, means that a large number of conventional power stations need to be kept part-loaded, or the national grid would become unmanageable. As a result, it would not be sensible for the UK to build more than 10 gigawatts of wind turbines.40
I have looked into his claims in some detail, and found that they cannot possibly be correct. The output of a single turbine – which might indeed be quite erratic – bears little relation to the output of all the turbines on an electricity grid, which fluctuates less rapidly. He has chosen the most sensitive portion of a wind turbine’s output: at greater or lesser windspeeds, a change of 1.5 metres per second makes much less difference to the power it produces. More importantly, wind forecasts made one day ahead have little bearing on balancing an electricity grid, which takes place over a much shorter timescale. In the UK, forecasts are made for the following hour, and the errors are generally very small.
The cost of electricity from wind farms, like the cost of all other forms of power, depends on who has performed the calculation: the question has become as much political as economic. The table below gives some estimates. They take into account all the costs of wind generation: buying the turbines, erecting them, connecting them to the grid and providing backup power. The average wholesale price of electricity from conventional power stations was 2.1 pence at the end of 2004 and 3.6 pence at the end of 2005.41
source |
onshore wind |
offshore wind |
Performance and Innovation Unit, 10 Downing Street – in 202042 |
1.5–2.5 pence |
2.0–3.0 pence |
International Energy Agency – present day43 |
3.0–7.0 US cents |
7.0–12.0 US cents |
|
(1.7–4.0 pence) |
(4.0–7.0 pence) |
Performance and Innovation Unit, 10 Downing Street – present day44 |
2.5–3.0 pence |
5.0–6.0 pence |
Sustainable Development Commission – present day45 |
3.2 pence |
5.5 pence |
4.8 pence |
6.3 pence |
|
Royal Academy of Engineering – present day47 |
5.4 pence |
7.2 pence |
i. It doesn’t say when.
The British prime minister’s office predicts that in 2020, electricity from gas will cost between 2.0 and 2.3 pence, while electricity from coal will cost 3.0–3.5 pence.48 These might be low estimates, as the wholesale prices have greatly increased since it published its report in 2002. Either way, if its figures are to be believed, onshore wind will provide the cheapest form of electricity in 2020, while offshore wind will be broadly competitive with conventional generation. The academics I have spoken to maintain that the Royal Academy’s figures are not supported by other studies and are considered by most analysts to be greatly inflated.
Though offshore turbines produce electricity more consistently than onshore machines, they are more expensive to build. Planting one in the seabed costs more than erecting it on land. The electronic equipment needs to be protected from the salt, and submarine cables are more expensive to lay than overhead power lines on land. But their costs will fall more rapidly than those of onshore windmills.49
Wave and tidal power are likely to remain more expensive than wind, as the table overleaf shows.
source |
onshore wind (per kilowatt hour) |
offshore wind (per kilowatt hour) |
Performance and Innovation Unit, 10 Downing Street – in 202050 |
3.0–6.0 pence |
|
Jake Chapman and Robert Gross – by 201251 |
4.0–5.0 pence |
4.0–5.0 pence |
Jake Chapman and Robert Gross – ‘short term’52 |
4.5–6.0 pence |
4.5–6.0 pence |
Performance and Innovation Unit, 10 Downing Street – ‘the first commercial-scale device’53 |
4.0–8.0 pence |
4.0–8.0 pence |
5.7 pence |
5.7 pence |
|
Royal Academy of Engineering – present day55 |
6.6 pence |
6.6 pence |
i. The Academy has not included the costs of standby generation for these technologies, because it does not expect much wave or tidal power to be used.
In one respect, the extra costs that renewable electricity might invoke have been exaggerated. The United Kingdom, like every developed nation, has a vast reserve of standby power that is, and will always be, maintained, and which is seldom included in the official figures: the emergency diesel generators owned by hospitals, army barracks, police stations, airports, offices and factories.56 They retain them in case the grid breaks down. Altogether there is somewhere between 12GW57 and 20GW58 of this invisible reserve: in other words, we have 16–26 per cent more generating capacity than the government’s numbers suggest.* If the higher figure is correct, and this reserve could all be called upon when renewable power failed, we would already have enough to support a 20 per cent contribution from wind, waves and tidal power of the kind described by Graham Sinden.59
National Grid Transco, the company which keeps the electricity on the wires, already uses a small amount of this hidden capacity. It has a ‘standing reserve’ agreement with a few of the people who own these generators, under which they must start contributing to the grid within 20 minutes of receiving its signal.60 This is quite easily achieved, as a diesel generator can reach full power from a cold start within 20 seconds. Wessex Water reports that it takes part in the scheme ‘primarily because experience has shown that standby generators won’t work reliably in an emergency unless they are full-load tested at least once a month’.61 There is, as far as I can see, no theoretical reason why the rest of the country’s emergency generators cannot be recruited for the same purpose. Because they already exist, and must continue to exist for other reasons, their use could greatly reduce the capital cost of standby power. The disadvantage of using these generators is that they have a lower fuel efficiency than large power stations, though this is offset somewhat by the fact that they respond faster. As they are likely to be used only for short periods, this use of fuel is not significant.
Another means of bringing down the cost is to make ambient energy more reliable. This sounds ridiculous: either the wind is blowing or it isn’t. But that is generally the case only in one place. The greater the number of regions in which windmills are built, the higher the chances that some of them will be turning. A study by the consultancy Oxera of fewer than half the possible wind-generating regions in the UK* discovered that in any one year there are, on average, only 23 hours in which electricity demand is high and wind turbines would be producing less than 10 per cent of their maximum output.62 Graham Sinden has studied the weather of the entire United Kingdom: ‘Between 1970 and 2003, there was not an hour, let alone a day or a week, with no wind across the UK.’63
When wind farms are 1,000 kilometres apart, their output is correlated at just 10 per cent: in other words, there is a 90 per cent chance that wind speeds will not be the same in both places.64
One of the advantages of high voltage DC cables is that they could reduce the amount of generating plant required to insure against renewable electricity failures, as they greatly expand the number of places in which wind and wave energy could be captured. Sinden’s work, as I’ve mentioned, also shows that renewable energy is more reliable if it does not come from just one source. The tides run whether or not the wind is blowing, and waves keep rolling across the sea long after it has dropped.†
But as the amount of renewable power increases, and its reliability improves, we encounter the opposite danger: that of an embarrassment of riches. If the wind is blowing strongly while demand is low, then turbines will have to be shut down if the frequency of the alternating current on the grid (assuming we are still using an AC grid) is not to rise beyond its limits. This seems like a painful waste of energy. But there may be a way of solving two problems at once.
At the moment there is only one means of reasonably efficient long-term electricity storage, and that is the process I mentioned in Chapter 5: pumping water from a low reservoir up to a high one and leaving it there until there’s a surge in demand. The energy losses involved in ‘pumped storage’ are about 20–25 per cent,66 which compare to losses of 60 per cent or more from the possible alternatives.*
The reservoir at Dinorwigin north Wales cost, at today’s prices, £1.6 billion to build.67 But it has a very long life, so the annual cost is actually quite low. If it were possible to build new pumped storage systems to soak up surplus power from renewable generators, their running costs would be small, as the electricity – which would otherwise have been wasted – is effectively free. They could also solve many of the problems associated with the intermittency of wind and waves and tides: when the wind drops, the gates can be opened. And in this case there would be no carbon costs.
So can more pumped storage plants be built? The House of Lords says that ‘the scope for increasing the volume of pumped hydro in the United Kingdom is limited by the same factors that limit conventional hydro.’68 But I cannot see how this could possibly be true. Conventional hydroelectricity is limited by the availability of rivers flowing rapidly downhill. But you don’t need a river to build a pumped storage plant: just a dip on the side of a mountain and another one at the bottom. By building small dams across both of them, you create a hydro system from scratch. It is easier still if one of the dips is already filled with water: a natural phenomenon commonly described as a lake.
The Royal Commission on Environmental Pollution raises a more plausible objection: ‘because of the effects… on landscape and wildlife, it is unlikely acceptable sites could be found.’69 This problem needs to be taken seriously. The dams and cables would certainly spoil the view. But given that many of our mountainsides are severely degraded by overgrazing, I am not convinced that the impact on wildlife will always be very great.
The attraction is obvious to anyone who studies a wind map of these islands. The greatest potential for generating power is off the coast of north-west Scotland. North-west Scotland happens to contain a large number of mountains, some of them close to the places in which the cables would come ashore. This means that scarcely any extra transmission networks would be required. But because there has so far been no public discussion of this idea, and no consideration of which mountains might be suitable and what the impacts would be, I am not putting it forward as a firm proposal. We should, however, start discussing it.
There is a further means of increasing the reliability of the system, and this is to alter the way in which people use electricity. If electricity use could somehow decrease when the wind drops, this would reduce the need for backup power, and the risk of the grid failing. But how on earth could that be done?
Several authors have suggested that appliances whose power does not need to be on all the time – such as fridges – should be designed to disconnect themselves when total demand rises.70,71,72 When the lights come on – on a winter evening for example – fridges and washing machines should turn themselves off. They could respond automatically to changes in the frequency of electricity. If the frequency dips (because there isn’t enough power on the grid), they switch off. If it rises, they turn themselves back on. This would reduce the great peaks in demand which force the electricity companies to keep so much plant on standby. It could also be used to help respond to the fluctuating levels of renewable energy. When demand is high and the wind is low, some of our appliances could switch themselves off. When demand is low and the wind is high, storage heaters, battery chargers and electrolysers could switch themselves on. This could answer one of the charges the critics of wind power make: that when the wind is blowing strongly, electricity will be wasted.
But there is a problem which most of the people who have written about ‘demand management’ have not addressed. The national grid company, sensibly enough, increases the frequency when it anticipates peaks in demand.73 It will do the same if it sees that supplies of renewable power are about to drop. So our future smart fridges or smart washing machines would turn themselves on just when they should be turning themselves off. But it’s not insuperable: the independent thinker Oliver Tickell seems to have solved it. Under his proposal – called the ‘Real-Time Pricing Initiative’ – the grid would double up as a communications network.
An open standard protocol [should] be developed and published to allow information about the instantaneous price of electricity to be broadcast through the electricity supply system. Much of the electromagnetic frequency spectrum would be available for this purpose without interfering with the principal power supply function, much as a telephone line can simultaneously provide voice and broadband internet service.74
‘Smart plugs’ attached to our fridges or washing machines would receive a signal that the marginal price of electricity is rising and switch their machines off, restoring power as the price falls back (or, in the case of fridges, before the temperature rises too much). He argues that
All the technologies involved are mature and available. All that is needed is good industrial design, and high production volumes.75
By altering the pattern of our demand, in other words, you can, in effect, improve the reliability of ambient power.
I am going to take four guesses, which appear to be supported by the evidence I have seen. The first is that the United Kingdom has sufficient renewable power comfortably to supply an average of 50 per cent of our electricity. The second is that the grid, and the reliability of the electricity it carries, could survive if 50 per cent of the supply came from renewables. The third is that the carbon costs of generating it would be considerably smaller than the carbon savings. The fourth is that the price per kilowatt hour would be no more than double the price the British government currently proposes for wind power supplying 20 per cent of our electricity.
In other words, my meta-guess is that 50 per cent is within the realm of feasibility. Depending on the cost of competing fuels, it might raise the price of electricity, but not grotesquely. I believe that the remaining 50 per cent, if we pursued a greatly accelerated development programme of the kind I discussed at the end of Chapter 5, could be supplied by thermal power plants whose carbon emissions are captured and stored. In other words, if my guesses are correct, all our electricity could be produced by two kinds of low-carbon generators: power stations burning gas whose exhausts are stripped of carbon dioxide, and renewable power plants, stationed either on our own soil or hundreds of kilometres away, and connected to the grid by means of long-distance cables. An electricity system running entirely on these two kinds of power (and conventional generators fired up to meet shortfalls in supply) would produce no more than 15 per cent of the carbon emissions currently released by our electricity suppliers. In combination with the efficiencies I discussed in Chapter 4, this would achieve an overall reduction of almost 90 per cent.
But I am sorry to say that, ambitious as this proposal is, it solves only part of the problem.
The reason is this: that 73 per cent of the energy we use in our homes powers our heating systems (for both space and water heating), and only some of this heat is supplied by electricity: 17 million of our 24.5 million homes have gas-fired boilers.76 Unless there is a means of solving the heating problem as well as the electricity problem, I will have to conclude that our homes – and our offices, factories, schools and hospitals – are unreformable. My targets, in that case, could not be met, and – given that we cannot do without heat – this project will have failed. I have somehow to find the means of heating our buildings without gas or coal. Because better insulation will give us a maximum likely carbon cut of around 40 per cent by 2030,77 at least 50 per cent of the carbon reduction will have to be found by changing the way we generate heat. Is this possible?
Heating, at present, accounts for about 24 per cent of our economy’s entire energy consumption.78 Altogether, we burn 2.4 exajoules of energy to produce the heat we need.79 An exajoule is a marvellous figure. It is a million million million (otherwise known as a quintillion) joules. Around 70 per cent of these exajoules are used in our homes.80
At first sight, the solution seems obvious: we should do what humans and proto-humans have been doing for a million years or so, and burn wood. Trees absorb carbon dioxide as they grow, so as long as you harvest them at or below the growth rate you will produce no more carbon dioxide than they consume. In Scandinavia, which has, so to speak, more trees than you could shake a stick at, this is a viable solution. Wood already provides 17 per cent of all the energy used in Sweden and 20 per cent of the energy used in Finland.81 But in the sparsely forested nations, it’s a rather tougher target.
In its thorough report Biomass as a Renewable Energy Source, the Royal Commission on Environmental Pollution estimates that the calorific value of wood is around 10 gigajoules per tonne, and that the most efficient means of producing it – growing willow trees and harvesting their branches every three years or so – produces about 10 dry tonnes per hectare per year.*82 Let us assume that growing, harvesting and transporting it uses 10 per cent of its energy content, and that the wood can be converted to useful heat at a rate of 75 per cent. Every hectare of land could then produce 67.5GJ of carbon-free heat. A gigajoule is a billionth of an exajoule. So to meet my target of 50 per cent of the heat we use – or 1.2 exajoules of zero-carbon energy – we would need 17.8 million hectares of land. The United Kingdom contains 17 million hectares of agricultural land, so we could just about accommodate it. You’ve probably spotted the flaw: we couldn’t grow anything else.
This only begins to describe the problem. Already, during the summer, large parts of the United Kingdom, especially its agricultural land, are becoming subject to water stress, partly because of over-abstraction and partly because of climate change. In another report, the Royal Commission notes that, ‘Because they are fast-growing, energy crops need more water than arable crops.’84 If we, and nations like ours, start devoting large areas of land to growing fuel rather than food, world food prices could start to rise, pushing malnourished people further towards starvation. In the longer term, the effect could be even more serious. In his book When The Rivers Run Dry, Fred Pearce has shown that falling water tables could threaten the world with famine.85 Energy crops like willow trees accelerate the process by two means: partly by ensuring that the water tables fall faster than before, and partly by keeping land out of food production as food becomes scarce. The danger is that having built the infrastructure associated with a major investment in fuel crops, we will be reluctant to switch back to food production when it becomes a moral necessity.
Growing wood for heating isn’t quite as pregnant with moral hazard as growing crops for road transport (which I will discuss in Chapter 8), for three reasons. The first is that woody plants can be raised on poorer land than the oil or sugar crops required by cars. The second is that growing wood for burning is a more efficient means of saving carbon dioxide. According to research conducted at Sheffield Hallam University, £1 spent producing bio-diesel saves between 3 and 6 kilograms of carbon, while one pound spent producing electricity from fast-growing trees saves 20 kilograms.86 Growing it for heat, where conversion efficiencies are higher, is likely to save even more. The third is that the value per hectare of wood production is lower than the value per hectare of oil crop production, so farmers growing wood would have a greater incentive to return to growing food if the world is faced with a shortage.
Even so, it seems to me that their contribution to water stress and the upward pressure on the food price means we should restrict the cultivation of energy crops to a maximum of 20 per cent of our land area. This would allow us to produce 0.23 exajoules of carbon-free heating, or 19 per cent of the 1.2 exajoules I’m seeking.
We could, of course, increase this volume by importing wood from more forested countries. But biomass takes up a great deal of space (a cubic metre of dry woodchips weighs 150 kilograms, while a cubic metre of coal weighs about 800 kilograms87) so it costs a lot to transport. Because its energy density per cubic metre is low, the carbon costs of trucking it can swiftly start to counteract the carbon savings of using it.
If it were burnt on the farm, the total emissions from using wood to produce heat would be just 5 or 10 per cent of those caused by burning fossil fuel.88 But for every 10 kilometres the fuel travels by road, the House of Lords calculates, 0.2 per cent of its energy value is consumed.89 If it travels 1,000 kilometres, the net energy saving falls by 20 per cent. This is a worse deal than the raw figures suggest, for you are swapping transport fuels – which are likely soon to become scarce – for heating fuels, which will remain abundant. Unfortunately, in places like northern Russia, where there are few restrictions on the destruction of forests, wood is much cheaper than it is in western Europe, so it makes economic – if not environmental – sense to truck or ship it over here. And its importation cannot currently be prevented. As the British minister for science and technology pointed out, ‘any restriction on fuel would not be permissible under international trade rules.’90
To the wood we grow, we could add the brashings, sneddings and sawdust produced by our foresters and the people who maintain our parks and gardens. This adds up, once other uses have been taken into account, to about 1.3 million dry tonnes a year,91 or another 0.7 per cent of my requirement.*
Our farmers produce almost 4 million tonnes of surplus straw every year.92 This will fall slightly if 20 per cent of our farmland is used for growing wood. The UN Food and Agriculture Organization, whose energy figures differ from the Royal Commission’s, estimates that wheat straw contains about 93 per cent of the energy per tonne in willow branches.93 For the sake of consistency, I’ll apply that percentage to the Commission’s figure, and give straw a value of 9.3 gigajoules per tonne. This provides 0.019 exajoules of carbon-free heat from 3 million tonnes of straw, or a further 1.6 per cent of the total.
The Energy Technology Support Unit also proposes that the United Kingdom could burn 1.3 million tonnes of chicken litter, which contain 13.5 gigajoules of energy per tonne, and about 1.8 million tonnes of ‘animal slurries’ (farmyard manure), whose energy content it does 120 not mention.94 But this suggestion disturbs me: it deprives the fields of manure, which has two major consequences. One is that more nitrogen fertilisers, which demand a good deal of fossil fuel to manufacture, will need to be produced; the other is that it will accelerate what could be an eventual global shortage of phosphate.95
So far, then, I have found 21.3 per cent of the 1.2 exajoules I am looking for. This assumes that the use of these biofuels is both straightforward and cost-effective. Neither assumption is entirely safe. If it turns out that the Food and Agriculture Organization’s figures are more reliable than the Royal Commission’s, we can raise this by a factor of 1.8, to 38 per cent. But you won’t be surprised to hear me say that I don’t know which numbers to trust.
The second most obvious source of carbon-free heat is the sun. The principle is simple: you place a panel of pipes resting on a black plate on your roof. The plate absorbs heat from the sun, warming the water in the pipes. The carbon costs of the system are approximately zero.
The consultancy company AEA Technology estimates that 50 per cent of the homes in the UK are ‘physically capable of accepting a solar water-heating system’.96 A report for the government by researchers at Imperial College suggests that if 50 per cent of our homes were fitted with solar heaters, they would produce 0.056 exajoules of heat, which is 4.7 per cent of our target figure. If the same ratio of heat production could be applied to our other buildings, the total contribution would be about 6.7 per cent. Unfortunately, solar hot water in the UK is, according to AEA Technology, ‘much more expensive than heat from fossil sources’, as the capital costs of installing the system are high, and the sun here is weak.
I am running out of options. Due to a regrettable absence of vulcanism, we have very few geothermal aquifers: bodies of hot water lying under the ground. The British Geological Survey says there are about 55 gigajoules of energy in aquifers with temperatures worth exploiting (40° or more). This means that, if it were not in the wrong place, you could mine it at the rate of 2.75 gigajoules per year for 20 years.97 This is a long-winded way of writing zero.
‘Ground-sourced heat pumps’ are more promising. Below 1.5 metres, the earth has a constant temperature of 12°. If you either drill a borehole under your house, or run a zigzag of pipes under the soil in your garden, you can draw the heat from the earth and concentrate it to about 50°. This is just about right for underfloor heating, though a bit too cool for radiators. As the heat pump works by circulating water through the pipes, you need an electric motor to drive it. But the system generates between 2.5 and 4 times as much energy as it uses.98 If the electricity comes from renewable sources, heat from the ground would be more or less carbon-free.
In the United Kingdom, we’ve been remarkably slow to use this technology. There are 230,000 ground-sourced heat pumps in Sweden and 600,000 in the United States.99 Here there’s a total of 300.100 One problem is that our housing stock is replaced so slowly. Sinking a borehole or digging trenches then laying pipes under the floor is more easily done when a house is being built than when it’s standing. Another is that we don’t have much space. AEA Technology assumes that ground-sourced heat pumps can be installed either in new homes or homes in the countryside which are not attached to the gas grid (of which there are 4.4 million). If heat pumps were installed in all the homes that could take them, they could provide a total of 79.3 terawatt hours of heat per year, which equates to 0.022 exajoules, or another 1.8 per cent of the number I’m chasing.*
In commercial or industrial buildings there’s the potential for a further 1.4 TWh per year, giving me another 0.033 per cent. I’m not doing very well.
All that remains is biogas: the methane produced by dumps, sewage farms and manure pits. Again, I am distressed to discover, the potential is small or less than small. AEA Technology says that constraints such as environmental hazard, public acceptance and the remote location of sewage farms (something for which in all other respects we are grateful) ‘will make it almost impossible’ for sewage gas to be used as a source of energy in buildings. ‘The effective heat market is almost certainly zero.’101 Gas from landfill sites (the term we now use to describe rubbish dumps) is subject to the same constraints: they are usually a long way from people’s homes, so it is expensive to pipe the heat to where it’s needed. On the whole, biogas is easier to use for generating electricity than generating heat.
And that, as far as I can tell, is everything unless we are greatly to increase our electricity supplies and persuade people to return to electrical resistance heating, which is slow, inefficient and expensive. Even if I use the most generous figures and ignore some of the constraints imposed by cost, I have located only 46.5 per cent of the 1.2 exajoules of heat I was seeking.
But before I give up and become an aromatherapist – which appears to be the dreadful fate of all disillusioned activists – I have one last throw of the dice.