Nuclear power is often cited as part of the solution of a low-carbon future, but detractors sometimes cite the CO2 pollution that would result from the concrete and steel needed for construction as a reason to hold back. While there may be other reasons for holding back, this particular one is spurious.
A way to measure the pollution performance of different energy sources is to divide the amount of CO2 pollution by the energy produced. For fossil fuels, looking at the fuel and not the plant cost (which we assume to be irrelevant), this number is between 300 and 900 g CO2/kWh, depending on whether you use gas, oil or coal, and how efficient your plant is. The CO2 involved with constructing (not operating or decommissioning, mind) is only 1 g CO2/kWh.
How much concrete and steel in a nuclear power station? The Nuclear Energy Institute claim (http://www.nei.org/index.asp?catnum=3&catid=1525) 400,000 cubic yards (306,000 cubic meters) of concrete and 60,000 tons (67,000 tonnes) of steel in a 1 GW rated nuclear power station. Let's work out how much CO2 this means, then divide by the energy generated over the productive life of the plant.
Notice our input figures are rough, so we're really only looking at one significant figure accuracy.
CO2 from concrete
How much CO2 is produced when making 520,000 cubic meters of concrete? That depends on the kind of concrete. There are different types and different figures. One way - take the density of concrete (2,300 kg/m^3 from the Physics Factbook), the CO2 to make cement (0.8 kg CO2/kg cement), the cement in concrete (10% from cement.org). This makes a figure of around 100 million kg of CO2 in our nuclear plant.
However, the Danish Technology Institute report (http://www.danishtechnology.dk/) is probably more authoritative. They claim that a cubic meter of concrete requires the production of 100 kg CO2, giving us 50 million kg of CO2 in our nuclear power station.
Comparing the figures, the order of magnitude matches. To be harsh, though, let's take the bigger figure - 100 million kg of CO2 for the concrete in a 1 GW power station.
CO2 from steel
How much CO2 is produced when 67,000 tonnes of steel is made?
Blue Scope Steel (http://csereport2005.bluescopesteel.com/) claim they put out 14.5 million tonnes of CO2 equivalent gasses in 2004/2005 to produce 5.72 million tonnes of steel product, which suggests around 2.5 kg CO2 per kg of steel.
Azom.com materials suggests around 2 tonnes of CO2 per tonne of steel, and Tata Steel claim (http://www.tatasteel.com/webzine/tatatech39/page14.htm) between 1.2 and 1.9 tonnes of CO2 per tonne of steel, depending on the process.
Let's be harsh again and pick 3 tonnes of CO2 for a tonne of steel. So we have another 200,000 tonnes of CO2 from the steel, or 200 million kg of CO2 from the steel to make a 1 GW nuclear power station.
Sum the steel and concrete CO2 figures: 300 million kg of CO2. If we had been conservative, that would have been 100 million kg CO2.
Energy from a 1 GW nuclear power station
If the power station produces power for a conservative 40 years, and runs for a pathetic 60% of the time (thus we're allowing for maintenance periods), the plant will deliver 210,000 million kWh of electricity.
The ratio is nearly zero
The simple ratio is 300,000,000 kg CO2 / 210,000,000,000 kWh - nearly 0.001 kg CO2 / kWh. Irrelevant.
This ignores the pollution from getting the fuel and running the plant. Also remember the CO2 is largely produced up front, which is bad news for quick CO2 reduction, but even building 10 GW of capacity to replace the UK's ageing plants will only produce 3 million tonnes of CO2 during construction - less than 1% of UK CO2 pollution in one year.