After nearly 20 years in the doldrums the nuclear power industry is gearing up for a revival on both sides of the Atlantic Ocean. The British and American governments seem likely to initiate new nuclear programs (pages 49-50), having listened to the industry's arguments that only it can provide substantial quantities of carbon-free baseload electricity while freeing the world from the price and supply uncertainties associated with gas.
The nuclear industry and its supporters further argue that nuclear generation costs are significantly cheaper than those of wind energy, that new reactor types will lead to lower construction costs -- and that this time the costs can be kept under control.
Much as the nuclear lobby would like us to believe that wind cannot provide a substantial proportion of electricity supplies, the claim is false. Operating power systems with substantial amounts of intermittent energy sources -- such as wind -- can be done, as independent and utility studies in the UK, Europe and the United States have been at pains to point out. In Denmark, system operator Eltra says integration of 50% wind by 2025 is technically and economically feasible. The UK system operator further makes the point that the issues are economic, rather than technical.
An electricity system powered in roughly equal proportions by wind, gas and coal is technically feasible. The decision on whether to pursue that route is part political and part economic. Rather than technical issues, the choice hinges on assessments of the necessity and cost of sequestering and storing carbon emissions from the burning of gas and coal, coupled with the cost of operating large quantities of wind power, with both set against the uncertain cost and potential hazards of operating nuclear power stations.
At first sight, the nuclear industry's estimates of its cost are compelling. Experts fuelling the fires of government policy makers tacitly concur that nuclear can deliver electricity at around $30-40/MWh, a price that wind cannot easily match. A new report feeding into the debate in Britain directly discriminates against wind, going so far as to baldly state: "From the point of view of the taxpayer, nuclear energy may be a strong contender given its costs relative to wind farms." It bases this conclusion on providing just 15% of UK electricity supplies from nuclear or wind, claiming that the extra costs of wind instead of gas generation are nearly four times the extra costs of adding nuclear instead of gas (page 47). In fact, given a level playing field, 15% wind on a system need cost the consumer no more than 15% of nuclear plant -- and could cost less.
On closer inspection, almost every presentation of nuclear generation costs uses financing assumptions appropriate to public sector projects, where governments take the risk of charging lower interest rates on long loans, knowing that if costs overrun or performance is not as expected, it can call on future taxpayers to keep a project alive. If the same assumptions were applied to large wind plant, hey presto, the price of wind is no more than nuclear. If, on the other hand, both nuclear and wind are financed in the private sector, with shorter periods in which to repay the capital at higher rates of interest, the tables are abruptly turned against nuclear and the generation cost of wind comes in cheaper (table page 43).
Financing has such a huge impact on generation cost for nuclear and wind because both have high capital and low running costs. As a result, the capital recovery period and the "cost of capital," often referred to as the test discount rate, significantly influence the generation cost. If the capital can be repaid over a long period of time -- 40 years or more is acceptable in the public sector -- at a low rate of interest, generation costs come down. To fairly compare wind and nuclear costs, both must be either treated as public sector projects, or private sector enterprises.
In the public sector
When it is the state, or state-funded utilities, taking the risk, capital recovery periods are often fixed at the lifetime of the plant. In the case of nuclear, 40 years is frequently used. The World Nuclear Association quotes prices based on a 60 year life, although no commercial reactor has yet to demonstrate that this is a realistic assumption. Test discount rates in the public sector vary between 3% and 8%. The combination of long life and low interest rates used in the prevailing assessments of nuclear costs means the capital recovery element of the generation cost is quite low.
The nuclear industry claims that by using new types of reactor it can construct power stations for much less than they cost in the 1980s. Evidence of this is in short supply, but industry estimates of installed cost roughly range between $1500/kW and $2200/kW for series build of nuclear power plants. The first stations, however, are estimated to cost up to 80% more than a bulk order from government might achieve.
Taking a cost in the middle of the range for series build of nuclear, a generation cost of $30.1/MWh results, made up of capital charges accounting for around $17/MWh, using the 5% discount rate and the 40 year life the International Energy Agency (IEA) employs for its nuclear estimates. The remainder of the cost is made up of operation and maintenance costs of $8.5/MWh, fuel costs of $4.6/MWh and decommissioning costs of around $0.3/MWh, also supplied by the IEA. Similar figures come from Finland, where a new nuclear project has recently been announced, and from most of the enthusiastic pronouncements from the nuclear lobby.
In comparison, wind's installed cost is a known, rather than estimated, $1000/kW. The wind does not blow at full strength all the time, however, so its load factor is lower than the 90-93% presumed for nuclear, which raises the cost of wind's capital. Based on load factors for wind in the 30-40% range, capital charges account for around $25/MWh, compared with nuclear's $17/MWh. With operation and maintenance costs for wind in the $8-12/MWh range, "public sector" wind energy generation comes in at $30-45/MWh, assuming lifetimes of 20-25 years and a cost of capital that is usually around 5% or 6%. In other words, cheapest public sector wind financed over 20 years at 5% interest is the same price as nuclear's best case estimate of $30.1/MWh, also at 5%.
In the PRIVATE sector
Investors in the private sector expect higher returns on their capital than governments and are not prepared to tie up their money for 40 or even 60 years. In wind power financing, project interest rates today rarely go below 8% (Windpower Monthly, May 2005) and range up to 12%, or more, depending on the perceived risks. Onshore wind is now perceived as relatively risk-free. The same cannot be said for nuclear. On both sides of the Atlantic Ocean, experts agree that investors want to see returns on investment in nuclear projects of 11-12%.
A private sector nuclear plant operator would almost certainly add a contingency to the installed cost to cover construction overruns, probably about 10%. The economic lifetime of a nuclear plant comes down to around 20 years in the eyes of private sector financiers, and the lifetime load factor drops from projection of 90-93% acceptable in the public sector, to 75-85% in the private sector.
The changes more than double nuclear's capital charges, raising that part of the total generation cost from $17/MWh to $40-50/MWh. Fuel and operation and maintenance (O&M) costs are less affected, although uncertainties push up nuclear O&M costs to $10-12/MWh. Decommissioning costs go up to $1/MWh, according to British Nuclear Fuels Ltd. Add it all together and the cost of electricity generated by nuclear plant financed in the private sector doubles, going up from the widely quoted $30-40/MWh derived using public sector financing assumptions, to $60-80/MWh. Some analysts have suggested the cost triples. These are the sums done by the private sector when it turned its back on the UK government's ambition to privatise Britain's nuclear power business.
Wind, as a technology with a deal more price-certainty attached to it than nuclear, does not suffer the same financial punishment at the hands of the private sector. Its cost and performance parameters do not change, although economic lifetimes come down to 15-20 years and so capital charges go up to around $40/MWh. The result is that total generation costs, on good sites in the United States or Europe, are in the $46-60/MWh range.
Generation cost crunch
On a level playing field in the public sector, wind and nuclear generation costs are similar. Moved from the public sector to the private sector, nuclear generation costs double, while those of wind increase by 50-60%. Wind, at $46-60/MWh, undercuts nuclear's private sector cost of $60-80/MWh by around 20%.
Nuclear's poor private sector economics provide an explanation for why the industry is appealing to governments for the softer financing terms of the public sector -- and would like governments to share the risks. The nearest approach today to a true private sector market for wind is in the US. It provides the evidence for the costs quoted here. Although British wind operates in the private sector, prices are distorted by the risks associated with the Renewables Obligation and the complex structure of the legislation, which by recycling the penalties imposed for not meeting the obligation holds prices up rather than pushing them down (Windpower Monthly, January 2004).
The nuclear industry's estimates for what nuclear power might cost under favourable public sector assumptions set against the price of wind power under Britain's Renewables Obligation is no fair basis for comparison -- even though that is the comparison that is frequently made. What consumers want to know is how much extra they will have to pay for unsubsidised wind power compared with unsubsidised nuclear.
The total cost
Generation cost is not the whole story. The cost of transmitting electricity to consumers and operating a power system needs to be paid for too. As long as wind or nuclear are providing no more than 10% of demand in the UK -- and assuming new wind and nuclear will be built instead of new gas -- adding wind is cheaper than adding nuclear (fig 1). For nuclear, this assumes the higher risk premium that financiers demand, but includes a generous load factor of 91% and only modest extra transmission costs. For wind it includes the extra costs of its variable supply.
As wind capacity rises, measures have to be taken to ensure wind variations do not reduce the reliability with which demand is met. Additional reserve plant need to be procured, to be brought into use when the uncertainties associated with wind power fluctuations exceed those due to demand fluctuations and power plant breakdowns. Taking a long term view -- and presuming no improvements in wind forecasting, system management or nuclear technology breakthroughs in the next ten to 15 years -- wind's need for extra reserve means that large volumes of wind will add more to overall system cost than large volumes of nuclear. The additional reserve can be part-loaded steam plant, open cycle gas turbines, storage, or demand-side management -- the market can be left to provide the most economic solution.
The costs are small, as are the extra emissions, because the reserves required are small. With 10% wind energy penetration, about $4 is added to a wind produced megawatt hour and with 30% wind, the extra cost is around $5/MWh. As more and more wind generation comes onto a system, however, its "capacity credit" for replacing other generation declines, a cost to the system that this analysis also adds to the price of wind.
Quantifying all the extra costs of wind power is a new science. Until now, the proportion of wind envisaged for power systems was so low that the extra system costs were not a concern. But with talk these days of adding 20% and more wind to power systems, the cost of high penetrations is becoming an issue. Recent studies in Ireland, Britain, Germany and Pennsylvania are shedding light on what the consumer will have to pay if wind energy is added instead of new coal or gas plant. The results of the studies are not directly comparable, as the underlying assumptions differ, but 10% wind adds around $2/MWh in the UK and $5/MWh in Germany.
Determining the extra cost of wind or nuclear also depends on the price of the new generation being displaced. In the UK today, it would be gas (fig 2) with a price tag of £33/MWh ($60/MWh), and expected to go on rising. The extra cost to the consumer for 20% wind instead of gas is a little over £2.2/MWh ($4.1/MWh) with gas at £0.25/therm ($4.8/million Btu), falling to £0.19/MWh ($0.36/MWh) with gas at £0.40/therm ($7.6/million Btu). For 30% wind instead of gas, the corresponding costs are £3.5/MWh ($6.5/MWh) and £0.44/MWh ($0.8/MWh), respectively.
HEAD TO HEAD
At low penetrations of wind energy into power systems, wind is undoubtedly cheaper than nuclear. Early nuclear stations are more expensive -- a fact the nuclear industry is clear on. With wind reaching levels of 20% and 30%, however, the differences become smaller, possibly tipping in favour of nuclear. Even so, the extra costs associated with accommodating significant amounts of wind energy into an electricity network instead of nuclear are modest.
Nuclear costs, however, are far less certain than those of wind. In the last wave of nuclear build in the 1980s, cost overruns were around four times greater than projected costs, according to the US Department of Energy. The projections being used today may still have large uncertainty margins. Furthermore, the nuclear costs quoted here are based on "series production" for nuclear plant -- an advantage not awarded to wind power. The British government has been told that the "first of a kind cost" for a nuclear plant would be 80% higher than the cost of the last plant in a production run of eight. It is under pressure to commit to an eight-reactor program -- a massive 10 GW of investment in projects that will take ten or 15 years to put in operation, if not longer.
Liability insurance is another nuclear issue -- and not one governments will like dealing with. Although the probability of catastrophic nuclear plant failure is low, the financial consequences of any significant releases of radioactive material could be extremely high. Any mathematician will confirm that the result of multiplying zero (the probability of failure) by infinity (the financial consequences of failure) is "indeterminate." That does not appeal to the insurance industry.