If the answer is nuclear you don’t understand the question.

When selecting technology to solve a problem, a common method employed is to have “eligibility criteria” and “merit criteria”. Eligibility criteria are the must haves; if you are selecting a new machine to generate electricity “does it generate electricity” is the obvious one. Eligibility criteria are all pass/fail tests. Merit criteria are what you use to split all of the options that pass. Say you’re comparing a diesel or natural gas engine to power a remote site, a merit criteria could be “is fuel available all year round?”

In this post I will focus on the eligibility criteria for powering Australia while reducing the greenhouse gas emissions associated with electricity production. I will spend a long time arguing that, no, nuclear should not be considered, because it is utterly incapable of solving the problem. In later posts I will discuss the merit criteria, where again nuclear falls down.

The public conversation, for want of a better term, about nuclear power has become mired in a feelings discussion about whether or not someone “supports” nuclear power, usually in some sort of hypothetical vacuum. This discussion is almost entirely without value; how anyone can be for or against a technology is beyond me. The question with any technology is how it is applied.

For this conversation to have any value it must be carefully framed. More “is nuclear power a good response to mitigating climate change in Australia?” than “is nuclear power nice?” or even “if we have nuclear reactors already built should they stay open?”.

Despite the events in Fukushima, which many observers consider a disaster, nuclear still gets a look in as a policy response to climate change. Most of the commentary surrounding the release of the White Paper focused on the favoritism or otherwise shown for nuclear power, and public advocates like Barry Brook still poke their head up  and shout “NUCLEAR” from time to time. While I will admit some of them have a point, anyone seriously advocating nuclear power as a policy response to global warming in Australia doesn’t understand the problem.

How does nuclear even get a mention in this discussion? There are two reasons in particular, and I admit they are quite compelling.

The first eligibility criterion for addressing electricity generation emissions is whether the technology generates less greenhouse gas than the technologies it will replace. This is determined using the full-cycle emissions intensity, which is the sum of all the emissions during production, mining, operation and decommissioning, divided by the generated electricity. It is hard to imagine any generation technology being truly zero emissions, as windmills require steel, solar panels require silicon and even hydro can release methane.

The figures below are Australian numbers, calculated by a team from the University of Sydney. In these comparisons it is vitally important to use region-specific data, as the emissions of construction are heavily driven by the emissions intensity of local electricity. Cement for example will usually be manufactured locally, and so the emissions of that process are governed by the grid emissions. Cement is a significant component of a nuclear plant, and so one drives the other.

Emissions intensity of a range of generating technologies. Figures in brackets indicate range. Taken from Lenzen, M (2008), linked from the image.

These figures were calculated before grid emissions and annual demand started trending down, which began in about 2009. The author notes that they assume that any new grid capacity will be best practice, which should drag the future emissions intensity broadly into line with current data. A motivated individual could go through the paper, determine the grid emission component of the energy use and then apply the curves, but it won’t be very much fun. It is sufficient to note that the emissions are likely to be lower than published, but not very much. This applies to all technologies, not just nuclear.

Over their lifespan nuclear power plants produce emissions per MWh somewhere between wind and solar, and about 10% of the emissions of a combined-cycle gas turbine, one of the most efficient conventional generators available. So a pass for nuclear on emissions intensity.

While this seems unintuitive, nuclear power plants are gigantic things made out of concrete and steel, a slow nuclear bomb in a box, while solar panels can go on the roof of your house, it starts to make sense once you analyse the other side of the metric. Nuke plants produce vast amounts of energy, due to the astonishingly high energy content of uranium.  The table below compares the energy content, in terms of electricity production, of a number of fuel sources. LEU at the bottom is enriched uranium, and the numbers are truly astounding.

Comparison of fuel energy content. The link through to the post that came from is worth a look, and includes an analysis of whether we have enough uranium to replace current energy sources.

This energy density sets it apart from coal, despite the generating technologies being very similar. Both plants make steam and drive a turbine to make electricity. Where a coal plant is very much a throughput problem, in the order of 100kg coal per second is required to keep an average plant running, nuclear plants require very little fuel, and the problem is getting the energy out of it slowly enough.

The second criterion, which might seem slightly ludicrous, is whether it is physically possible to make enough electricity from the technology. These calculations are most often applied to biofuels options, but are equally valid for electricity generators.

The high energy density of uranium means that the technology is quite space efficient, a key advantage over current renewables. As a coarse estimate, we could replace all the coal generation in Australia with 25 or so plants, and provide all of our energy with 50ish, depending on what you include and grid requirements for peaking power. To do the same job with wind would be more in the order of 50,000 – 75,000 turbines, or a square 100km by 100km with solar panels.

So, I concede, when considering the hypothetical, technology assessment of nuclear, it sounds useful. Loads of energy, low emissions. But how appropriate is nuclear power as a response to climate change?

The warnings we routinely ignore from the International Panel on Climate Change are unequivocal; everyone must reduce their CO2 emissions by 40% by 2020 and by +90% by 2050. Nuclear power in Australia should only be considered within this context, because if you don’t care about climate change then we might as well burn all the coal we have. And we have a lot.

The current policy combination of the Renewable Energy Target and the carbon price seems to be working, although probably not quickly enough to hit those targets.

What happens then if we decide on the 1st of January 2013 that we are going to go nuclear and start construction on the same day?

There are two things to consider when analysing the impact on emissions a new build can contribute; how quickly can you build it and how long will it take to pay back the emissions from the construction phase?

Build times vary, depending on your source, but there are a few builds underway to provide some signposts.

In 2009 the UK government decided to build 10 plants. That is just actual power plants, they already have an established nuclear power industry and will be able to leverage some skills and infrastructure that Australia doesn’t have. I’m having trouble finding references on construction time, but this is indicative “Ministers hope to fast-track the construction of the new plants so that some can be producing energy by as early as 2018.” Nine years from “yes we’re doing this” to electricity. Other similar builds in Georgia (US), France, China and Finland are on track for about 8-9 years as well.

Electricity production would commence in 2022 in the Australian scenario. Two years after our emissions should be 40% less.

Any low emissions technology now has to start saving tonnes of greenhouse gas to earn back the construction emissions. It won’t be making a positive contribution to greenhouse gas mitigation until it has saved more tonnes than it cost to build.

The whole of life cycle comparison between construction emissions and tonnes saved is incredibly complex and thus susceptible to a large degree of error. How do you account if the contruction emissions are 90% diesel, but it only offset electricity use? Do you include projections for average emissions intensity in the offset of electricity? This whole difficult metric is often referred to as the carbon payback, and is the sort of thing that requires a thesis to calculate properly. Some numbers are available, wind here with a payback of about 3.5 years,

A reasonable proxy then is the simple energy cost of production against the time it takes to generate that much electricity. Often called energy payback.

As emissions have been removed from this calculation regional variance is less significant. However, Lenzen (2008) linked above, has calculations for Australian conditions. Page 7 here makes it pretty clear:

The energy payback time of nuclear energy is around 6½ years for light water reactors, and 7 years for heavy water reactors, ranging within 5.6-14.1 years, and 6.4-12.4 years, respectively.

So if we start construction at the beginning of 2013, nuclear power will start reducing Australia’s emissions from electricity production in about 2028.

Any comparison with wind or solar will need a detailed list of assumptions, but reporting a range is close enough. The graphic below plots “energy return on investment” for various sources, which is an analog for energy payback. It is the ratio of electricity generated over the lifetime of the plant and the energy used in construction and operation. Higher numbers mean shorter payback.

The Offshore Valuation: A valuation of the UK’s offshore renewable energy resource. Published in the United Kingdom 2010 by the Public Interest Research Centre. ISBN 978-0-9503648-8-9 http://www.offshorevaluation.org PE (2006)

I haven’t seen Australian figures for this and would be interested to see them done, but other sources support the values above. Converting these to energy payback puts wind at 6 months to 2 years and solar at 1.5 to 5 years. As these are renewables their output will of course depend on their location.

It should be clear by now that if your goal is to reduce emissions as quickly as possible the best way to achieve that is using renewables. Nuclear power actually cannot contribute in the next 15 years. And that’s from the start of construction. One expects that overcoming the public opposition to nuclear, and getting it onto the political agenda will take years on its own.

Advocating for nuclear power in Australia is advocating to do nothing about our emissions until almost 2030. Wind, solar and energy efficiency are reducing our emissions right now and will continue to do so.

The very reasonable point has been raised elsewhere that the above does not address the long term scenario; could nuclear contribute substantially in the 2050 scenario?

The incredibly thorough UMPNER review, released in 2007 doesn’t offer much hope:

The greenhouse gas emission reductions from nuclear power could reach 8 to 17 per cent of national emissions in 2050.

So if we abandoned the 2020 target of a 40% reduction, went nuclear and held out hope for hitting the 90% by 2050 target, we would also fail fairly spectacularly. Further, if we miss the 2020 target we also need to offset those additional emissions in later years.

So if your answer is nuclear, you don’t really understand the question.