Within its five-year life span, this government is going to have to make sure that some difficult and potentially unpopular long-term infrastructure decisions are made.
The trickiest of these relate to power generation. In the muscular days of the Central Electricity Generating Board (CEGB), the minister of power (remember Manny Shinwell?) could instruct the board to do what he wanted. That all changed with privatisation in 1989-90. The CEGB vanished, and the role of government changed to that of facilitator and regulator. Yet were the lights went out, the government would be blamed, even though it can no longer order the building of power stations. Tony Blair's statement at the Labour party conference, promising a fresh energy review early in 2006, and accepting that nuclear energy will have to be part of it, indicates that the government is aware of its predicament.
Supply and demand for electricity
At the moment there is a surplus of generating capacity. The UK has a total capacity of about 77GW (Didcot power station can generate about 2GW) compared with the record peak demand of 61.7GW that occurred at 5.30pm on 10th December 2002. Electricity demand in Britain is growing only slowly so, allowing for a safety margin for breakdowns, surplus capacity ought to exist for at least 20 years. Unfortunately that is not the case.
We have 174 non-nuclear power stations, with a total capacity of 62GW. Of these, 52 per cent in power capacity terms are more than 30 years old. Over the next 20 years most of this elderly capacity will have to be replaced. In addition, all but one of our 12 nuclear stations will have been retired by 2025. Replacements amounting to more than 40GW will be needed—plus additional capacity to meet growing demand. Since the transmission grid works better with a scattering of smaller stations rather than a few whoppers, a minimum of 30 new stations will be needed.
Because of the time it takes to build new stations, the government has to persuade the generation companies to make the major building decisions in the next five years and work must start soon on at least 20GW of new capacity. Of the more than 50 new stations in at least the early stage of planning, most are wind farms of small capacity. So far only four substantial new stations, all gas-fired and totalling 3.1GW, are likely to come on stream by 2010.
The national grid's forecast based on current trends envisages an annual growth in demand of about 0.8 per cent. Many things could change—high prices might reduce or even reverse the trend, growth in air-conditioning might increase it, and so forth. (What the forecast does not consider is the big increase in electricity demand if road fuel shifts to hydrogen or rechargeable batteries. To give a feel for the magnitude, if all of today's petrol was replaced with hydrogen made by electrolysis, Britain would need the equivalent of 30 Sizewell B nuclear plants.)
Since privatisation, our energy markets have been intensely competitive—more so than almost anywhere else in the world. Unsurprisingly, when that competition was combined with the surplus generating capacity, the wholesale electricity price fell sharply. By early 2003 it had fallen to £15 per megawatt-hour (MWh), a price that made it impossible for power companies to contemplate investing in new capacity. Indeed, British Energy and the British subsidiaries of two American generators, TXU and AES, went bust. But prices have since rebounded, and the April 2005 futures market price reached a record high.
It is a central tenet of market economics that prices provide the signals needed for accurate decision-making. For most goods and services this is indeed the case. But it is not always true. The electricity futures market covers only the next three years. Right now the 2006 market may be offering the highest price ever recorded, but this is little comfort if you are planning a new station that will not come on stream for another ten years. The marketplace does not send reliable signals when investments need long lead-times.
The power generation industry is not unique. It shares characteristics with a small number of other ventures—oil refineries for instance—that render long-term investment decisions particularly tricky, and market signals unhelpful. These characteristics include large front-end costs, long lead-times before profits begin to flow and a highly competitive market. This is why, for example, it has been almost impossible to justify building a new oil refinery for almost 30 years. Today's shortage of diesel fuel and inadequate capacity for refining crude oils is the result, but at least with oil refining, output can be stored and used later to ease shortages. Not so with electricity.
The uncertainties facing a generator A generating company thinking of making an investment in capacity faces daunting uncertainties. In addition to inadequate long-term market signals, there are also regulatory, environmental and technological uncertainties.
Regulatory. The government has created a regulatory agency, Ofgem, to ensure that the industry operates in the public interest.
In its 2003 white paper, the government put great emphasis on renewable energy—such as wind power, tidal power and burning biomass. The regulator can encourage renewable energy through the renewable obligation certificate (ROC). Any generator of 1 MWh of renewable electricity is given one ROC by Ofgem. It can sell this to suppliers who have, by law, to demonstrate that 5 per cent of the electricity (increasing to 15 per cent by 2015) they buy from generators is renewable. So the generator of a MWh of renewable electricity has two sources of income—sale of the electricity (about £24 at the moment) plus sale of the ROC (currently fetching about £48) for a grand total of £72—about three times the actual value of the electricity. It is a neat arrangement, providing a hefty subsidy to wind farms and other renewable sources, without falling foul of EU subsidy rules.
The idea could be extended to nuclear obligation certificates, carbon storage obligation certificates or whatever the DTI fancies. But for a generating company wanting to invest, the uncertainty surrounding the availability, magnitude and value of future obligation certificates adds to the difficulty of decision-making.
Environmental. Now that the Kyoto protocol is in force, the environment ministry, Defra, is responsible for running the carbon dioxide trading system. The unit of trade is the assigned amount unit (AAU). Defra gives AAUs to power stations, oil refineries, cement works and other plants that emit large quantities of CO2 on the basis of their past record. When a plant emits one tonne of CO2 it eats up one AAU. If it runs out of AAUs it must buy more or pay a fine of E40 per tonne of CO2. If it finds ways to reduce its CO2 emissions it can sell its surplus AAUs.
This trading system is also subject to great uncertainty: how many AAUs will be granted to a plant in future, the cost of a fine, and the market price of an AAU are all unknown. Much will depend on how serious we are really going to be about global warming. Each has an impact on any decision to invest in new generating capacity, or to modify existing plants to reduce emission levels.
Technological. A company contemplating building additional generating capacity will also have to decide which technology to use. The table below lists the major contenders, with the full-cycle (a power station's lifetime) cost and the full-cycle CO2 emissions shown for each. For costs, the figure shown includes everything from initial planning through construction, operation and decommissioning—from greenfield back to brownfield. For emissions, the figure covers the CO2 emissions from the steel, cement, transport and other inputs needed to build and then decommission the plant; the CO2 emitted when mining, processing and transporting the coal, uranium, biomass or other fuel; and the CO2 emitted when generating the electricity and (in the case of clean coal) storing the CO2. (These figures depend on many assumptions, which partisans slant the way they prefer, but they provide as fair a picture as is possible.)
Full-cycle costs and emissions for the main technologies
| Technology | Cost | Emissions |
| (£ per MWh) | (kg of CO2 per MWh) | |
| Existing coal | 25 | 1,000 |
| Optimum new coal with CCS | 50 | 20 |
| Optimum new gas | 22 | 430 |
| Onshore wind | 37 | 14 |
| Offshore wind | 55 | 14 |
| Nuclear fission | 23 | 15 |
Coal, gas, wind and nuclear Important though they are, costs and prices are not the only considerations. Safety and security are important too. With this in mind, it's worth looking at each fuel in more detail.
Coal. Coal is abundant. Worldwide reserves amount to about 200 years at current consumption rates. When probable and yet-to-find deposits are added, reserves may be more than 300 years. World coal reserves are widely distributed and the major exporting countries (Australia, South Africa, Indonesia and the US) are not obvious partners for a cartel, so security of supply should continue.
The disadvantage of coal is its high carbon dioxide emission per unit of energy output, as we can see from the table. It is unlikely that any generating company will ever build a conventional coal-fired power station in Britain again. However, combined with the capture and storage of the resultant CO2 (called CCS) coal-fired stations represent a promising option for the future (see Prospect, March 2003).
Gas. Everybody loves gas. Gas is plentiful at the moment. It is quick and easy to build gas-fired power stations and gas is relatively clean, emitting only about half the CO2 emitted by coal. It is easy to forget that EU rules once forbade the burning of this "premium fuel" for mere power generation. That rule had long gone by the time electricity was privatised in Britain, and the subsequent "dash for gas" in the early 1990s delivered the final blow to Britain's coal and nuclear industries.
Unfortunately gas production from our own fields is now in irreversible decline. Increasingly, we will have to import gas, so security of supply becomes an important issue. Foreign gas arrives in this country by pipeline and also in liquid form (LNG) by tanker. Gas from friendly Norway is one thing, gas from "the Stans" and other distant and unstable places is quite another. The image of Chechen rebels blowing up a remote but vital pipeline bringing gas to Britain makes for good television drama, but it also serves as a warning. With multiple gas sources, a network of pipelines and extensive use of LNG, we can greatly reduce those security risks.
The size of the world's total gas reserve is very uncertain. The figure for proven plus probable plus yet-to-find gas is thought to be about 100 years at current world consumption rates. This estimate will shrink rapidly if world demand grows as fast as many people expect. Any prudent generating company planning a large gas-fired power station with a design life of 20 years or more must surely consider resource limitation very carefully.
As with coal, a generator building a gas-fired station would be wise to allow for the cost of carbon capture and storage if stringent targets are imposed by Defra—as they will be.
(Oil should be mentioned in passing, if only to dismiss it. Little electricity is generated using oil as a fuel, although diesel-electric sets can be important locally. In any event, world oil production is widely expected to start declining before the end of this decade, leading to a sustained increase in oil prices.)
Wind. Rising levels of public objection to unsightly onshore wind farms will make it more attractive for the investor to build his wind farm offshore. With today's wholesale price and ROCs worth £48, offshore wind farms are economically viable—hence the numerous plans that are being announced.
Because wind blows intermittently, the typical windmill yields about 30 per cent of its capacity. A few are so poorly located that they yield less than 20 per cent, and a handful yield more than 50 per cent. It used to be widely believed that wind farms would require a lot of gas-powered (and CO2 emitting) backup to maintain a steady load. In practice that is turning out to be less necessary (and costly) than feared. To see why, we need to remember that the grid—the system operator—is primarily concerned with the balance of supply and demand from minute to minute; the "co-ordinated kettle" demand surge at the end of a popular television programme upsets the balance far more than wind fluctuation.
As long as wind power does not exceed the white paper's target of 20 per cent of our total generating capacity (6 per cent of total production after allowing for intermittency), only a small amount of back-up generating capacity is likely to be needed. At present the 1,237 windmills both on and offshore make up 1.3 per cent of the nation's installed capacity, so another 15,000 or so windmills can be built before back-up becomes a significant issue.
Other renewables (wave, tidal, biomass, landfill gas) are unlikely to provide generators with significant enough opportunities.
Nuclear. For the generating company, nuclear fission offers an approach tantalisingly full of promise, yet surrounded by formidable obstacles—real or imagined. Three elements can be used to fuel nuclear reactors—uranium, thorium and plutonium. The first two are naturally occurring and the third is man-made, a by-product of nuclear reactors.
In a reactor, some uranium transmutes into plutonium; indeed, one type (the so-called "fast-breeder" reactor) produces more fuel than it consumes. In principle, the spent uranium fuel rods can be reprocessed to recover this plutonium. This makes sense if uranium becomes scarce or expensive, but it has the security disadvantage of making the plutonium much more accessible. Spent fuel rods are intensely radioactive but, once separated by reprocessing, plutonium is not nearly as radioactive and small amounts can be carried without much personal danger for short periods. Given today's concerns about terrorism, many people see reprocessing as undesirable.
Thorium is more expensive than uranium as a fuel, which is why there are no thorium-fuelled plants. There is, however, no technical reason why thorium could not be used.
At the most basic level, known uranium and plutonium reserves, which include surplus military material, are sufficient for about 50 years at the present rate of consumption. If there was a rush to nuclear, this reserve life would shrink. However, geological exploration will almost certainly uncover more uranium. If this is the case then fast-breeding, reprocessing and thorium will not be needed. Nevertheless, it is comforting to know that these three technologies are available, just in case. Even at much greater consumption rates we can be fairly confident that reserve life extends beyond 100 years.
Disposal of nuclear waste is "the issue that corrodes the entire picture," as one pro-nuclear expert put it recently. The technical aspects of disposal pose few remaining problems; it is the conflict between the ethical and the political issues that is so intractable. For the last 50 years we have taken the easy option: even the most dangerous waste has been stored in temporary surface containers. But failing to clear up our own mess as we go along is little short of nuclear child abuse. On the other hand, no politician relishes the task of convincing the electorate to accept a permanent underground waste repository in this green and pleasant land.
This conflict must be resolved before any more nuclear fission power stations are built. Shareholders would roast any company that proposed building a new station without a clear idea of the nature and cost of the waste disposal obligation involved—unless the state offered to carry that responsibility.
Despite the three big nuclear incidents of the last 50 years (the accidents at Windscale and Three Mile Island, and the Chernobyl disaster, which has caused 56 deaths over the 19 years since it occurred), modern reactors are very, very safe. Unfortunately, comforting statements of this kind are difficult to prove and not widely believed. Getting local opinion to accept a nuclear plant in its backyard (unless it already has one there) remains an enormous hurdle.
The cost of a nuclear fission plant—including decommissioning—is the least of a generator's problems. Reference to the cost of Britain's existing plants does not provide a useful guide—a mixture of military imperative and technical hubris made our past programmes excessively expensive. The Finnish and Swiss experiences provide a more reliable cost guide, and form the basis for the surprisingly low figure included in the table.
It is doubtful whether we in Britain have the technical capability to design, build and commission nuclear plants any longer but, once we have swallowed our pride, we can probably import the necessary expertise and skills from France or America.
How can government guide decisions? The days when a government could impose its will on a nationalised generating industry are gone. Now, all it can do is take steps to ensure that the economic and regulatory climate is benign enough to attract investors. The risks of technology, fuel reserves and electricity price are ones that the generating industry can reasonably be expected to shoulder. But the costs of global warming, and the particular risks associated with nuclear fission will need to be shared between the state and the industry to ensure the level of generating investment that we need.
Emission risk. New Labour's much trumpeted aspiration to achieve a 60 per cent reduction in CO2 emissions by 2050 is too woolly to help a company planning a new power station. It needs, first, an explicit long-term emission-reduction goal that has all-party political support, together with obligatory "way-points" that must be met en route to that goal, and a rule as to who will pay what penalties if those way-points are not met.
Second, a generating company needs a clear statement as to who is going to pay for that emissions commitment. It can only be the consumer or the taxpayer: any attempt to make the generator pay will simply lead to a refusal to invest. Making consumers pay has two advantages—it hits the people ultimately responsible for making the mess, and payment mechanisms already exist (ROCs, AAUs) that can be elaborated as needed. It has the disadvantage of heaping the costs on to electricity-consuming industries, undermining their competitiveness with those in countries that do not curb emissions, such as the US. For domestic consumers, such a payment strategy would be seriously regressive.
Making the taxpayer pay avoids the competition and regression problems, but brings other problems. Establishing a verifiable and cheat-proof system whereby the government made tax-funded purchases in the emissions market would be a far from trivial task. An intriguing possibility would be for the DTI to tender to buy stored CO2—initially inviting bids to supply, say, ten packages, each of a million tonnes a year over 20 years, commencing in, say, 2020 (to give a long enough lead time). It might also give Britain pole position in the race to market carbon capture technology—a theme running through the 2003 white paper.
Obviously no taxpayer is going to welcome a new tax: yet this might be one of the few occasions when a fully hypothecated tax would make sense and increase its political palatability. The amount involved, about £15 billion a year, looks large but in terms of our trillion pound economy is fairly modest. Add to this the effect of ROCs and the almost inevitable steep increase in petrol and diesel prices, and the government's 60 per cent cut in CO2 emissions starts to look achievable.
Nuclear risk. The open-ended nature of the obligation to keep nuclear waste secure in the short term from terrorists and in a long term measured in centuries, makes it inevitable that the state will have to police and pay for these tasks. An explicit recognition of this—perhaps in the form of waste security "put options" that could be bought by generators from the state—would remove one of the impediments to future nuclear investment.
Portfolio risk. Each generating company will make the optimal technology decision for its own new power station yet as a country we may end up without sufficient diversity. Unfortunately market signals, weak and easy to misinterpret for these particular investment decisions, may hinder rather than help—encouraging a dash for this or that technology.
Conclusion After privatising power generation, government has few sticks, and its carrots have to be extra juicy. The juiciest carrot a government can offer a potential investor is reduction in uncertainty and willingness to share risk.
Wind power alone is free from reserve risk. Gas has potentially worrying reserve limitations—particularly if the whole world chases it. Nuclear has its own well known problems. Coal (with carbon capture and storage) is less constrained in these respects and a small but growing number of specialists, myself included, believe that CCS, not nuclear, is probably the wisest choice for at least half of our future power needs.
Back in the days of the CEGB, politicians used their authority to "pick winners" (all too often turning out to be losers), rather than to ensure diversity. Today, the decision as to which technology to use rests with the investing companies—let's hope they don't all dash in the same direction.
