David MacKay’s foreword to COP21 edition of Climate Gamble

Professor Sir David J C MacKay, famous for his excellent book “Sustainable Energy: Without hot air,”  very kindly provided the following foreword for our COP21 edition of Climate Gamble. He passed away on 14th April 2016 and will be much missed by everyone interested in “pro-arithmetic” energy discussion.

Climate change action is remarkably difficult.

Society has many levers available:

  • demand-reduction through lifestyle change or technology changes;
  • eating less meat;
  • bioenergy;
  • wind power;
  • solar power;
  • hydro-electricity;
  • carbon capture and storage;
  • nuclear power;
  • carbon-dioxide removal;
  • reforestation;
  • solar radiation management;
  • population reduction.

Every lever has technical limits and political difficulties. Bioenergy, for example, requires very large land areas, and may have environmental impacts. Eating less meat could make an enormous impact [see globalcalculator.org], but many view vegetarianism as a political non-starter. Some people object to the land area required for wind power and solar power, and the intermittency of wind and solar is a technical challenge.

Making a plan that adds up and that is politically and economically credible is not easy.

Anyone who suggests that one of these levers should not be used by society must recognise that this constraint inevitably makes the task of climate change action harder.

I think that some people view nuclear power as untouchable because the language for describing the dangers of nuclear radiation is too black and white. When we talk about other forms of radiation, everyone understands that there is a scale ranging from harmful to harmless, and we have nuanced language to distinguish between, for example “desert sunshine” and “moonlight”, and other levels of “bright” or “dim” radiation in-between.

Everyone knows that midday desert sun can be harmful if one lies in it without protection. And everyone knows that moonlight is essentially harmless. Yet moonlight is made up of just the same photons as sunshine! The reason why moonlight is harmless is that it is 400,000 times less bright than sunshine.

When people talk about nuclear radiation, our language lacks analogous terms for “bright” and “dim”. Nuclear radiation is just said to be “toxic”, “harmful” or “dangerous”. Black and white. But in fact nuclear radiation can be like sunlight, and it can be like moonlight. There are levels of radiation that are lethal, and levels of radiation that are essentially harmless.

Responsible citizens should not simply rule out nuclear power from the portfolio of climate-change options without properly, quantitatively understanding the true risks.

Yes, the nuclear industry has had accidents. Yes, in some countries, the nuclear industry has had a reprehensible track-record of mis-management and dishonesty. Yes, nuclear waste lasts a long time (as do many environmental pollutants).

But please don’t leap to simplistic conclusions.

We owe it to our children to behave like adults.

Have an open mind.

Read. Learn. Think.


David J C MacKay FRS

Regius Professor of Engineering

Cambridge University Engineering Department

Author of “Sustainable Energy – without the hot air” 

and “Information Theory, Inference, and Learning Algorithms”


Former Chief Scientific Advisor, Department of Energy and Climate Change, UK.

Dangers, or external costs, of different energy sources (Weekly pic)

External costs of different electricity generation technologies in the European Union (2012). Adapted from Ecofys (2014): Subsidies and costs of EU energy, figure 3-8 and Annex 1-3, Table A3-8, p. 109.

How dangerous the different energy sources actually are? Evaluating the problems of different energy sources is always wrought with controversy, as there is not and can not be a single, universally acceptable method of evaluating the severity of what may be very different impacts. For example, how one should evaluate health effects on children compared to elderly, or health effects on humans to damages to environment?

Nevertheless, some commendable attempts at comparing the dangers of energy technologies have been made. Among the more recent ones is a study commissioned by the European Union and performed by consulting company Ecofys. Known for its pro-renewable and anti-nuclear reports, Ecofys estimated the average so-called “external” costs of different electricity generation technologies in use within the EU. In this context, external costs refer to costs the energy generators do not have to pay, but instead “externalise” to the society as a whole.

The Ecofys study referenced here is not without its problems, but it nevertheless remains an admirable effort at evaluating the dangers of different energy sources on a level playing field. As we can see from the results, even the highest estimate for the dangers and costs of nuclear power comes below the costs of natural gas, and comfortably close to the costs of biomass use – whose radical expansion is one of the keystones of every non-nuclear energy scenario presented by traditional environmental organisations.

The low estimate, on the other hand, comes well below the impacts of solar power. That’s even when the impact of nuclear accidents is factored in.

If we take the Ecofys study at face value, we should probably conclude that the dangers of nuclear energy are on par with dangers associated with most forms of renewable energy. The dangers may be different, of course: renewable energy sources do not experience meltdowns, but they may emit considerable pollution during their construction or use. The end results, however, are similar: both acute and chronic illnesses to people exposed.

It should be noted that the findings of this Ecofys study are in line with findings of a review commissioned by Friends of the Earth UK: in the review of relevant scientific literature, the report concluded that

“Overall the safety risks associated with nuclear power appear to be more in line with lifecycle impacts from renewable energy technologies, and significantly lower than for coal and natural gas per MWh of supplied energy.”

Carbon intensity of electric power sources (Weekly pic)

IPCC (2014) median estimates of life cycle carbon intensity of selected electricity sources. The figure includes mining, raw material and waste disposal impacts, but excludes infrastructure requirements such as energy storage, strengthened transmission grid, or backup generators. As a result, the figures are likely to underestimate emissions from variable sources such as wind and solar power.

To avert dangerous climate change, we will need lots and lots of low-carbon energy sources. Electricity is perhaps the most important of these, as it is wonderfully flexible form of energy that can replace fossil fuels in multiple applications. Furthermore, it is easy to deliver and we know how to generate it in quantity with very low carbon emissions.

Of all the methods of low-carbon electricity generation, nuclear power is still the single most important. It alone produces far more low-carbon electricity than all the “new” renewables combined. This is an inconvenient fact for those who try to oppose nuclear power while simultaneously opposing climate change. As a result, one hears constantly claims that nuclear power produces greenhouse gases – and that this makes it unsuitable for climate mitigation.

The first part of this claim is true: In fact, no energy source produces energy without greenhouse gas emissions of any kind. There are emissions associated with wind power, and there are emissions associated with solar power. The second part, the inference, is false, however. In all the serious research on the subject, the carbon balance of nuclear electricity is found to be very low. It compares well with wind power and, in fact, tends to be lower than that of solar electricity.

The most common counterargument we’ve heard at this point is the obvious: “Wait, what about uranium mining, or building of nuclear power plants? Surely they contribute a lot of emissions?” 

Fortunately, we can say that this counterargument does not hold water. The figures quoted here, and in any serious scientific report, are so-called lifecycle emissions. This means that the figures already include impacts from mining, building of power plants, and so forth. In our opinion, it is somewhat insulting even to think that such obvious emission sources would not be included in any serious calculations.

But one thing that’s not included in these calculations is the additional infrastructure that is required to deliver equivalent level of service. For nuclear power plants, not much additional infrastructure is required, beyond obvious power lines. But if we want to deliver equivalent service – equivalent amount of reliable electricity generation – from variable sources, we most probably are going to need more infrastructure. This can be reinforced electricity grid to transfer energy from places where the wind blows or the sun shines to places where electricity is needed; it can be backup generators that provide power when it’s dark or calm; or it can be wind turbines and solar panels that are part of “overbuild” required to ensure that at least some catch the wind or the sun at all times.

So far, as the share of variable energy sources in electricity grid has remained small, this additional infrastructure is not really needed. The existing grid and existing power plants can cope with the limited variability, although this often incurs extra costs already. But as we expand our low-carbon energy production, we will need more and more infrastructure to cope with increasing variability. This causes both economic and environmental costs, resulting to higher carbon balance than these simple calculations would suggest.

This is one of the reasons why we believe that opposing proven solutions that can provide significant quantities of low-carbon energy is, at this point, a gamble with the climate. For more information, buy our book, Climate Gamble – or come to Paris during the COP21 climate negotiations and get one for free!


How fast can nuclear be built? (Weekly pic)

Greatest increases in low-carbon electricity generation over 15-year period, adjusted to account for population differences between countries. Data from BP World Energy Outlook 2014.

Even at this late hour, when we have mere 35 years to effectively end the burning of stuff for energy, there are some who argue against low-carbon energy generation. One of the more common arguments is that nuclear power plants take too long to build, and therefore they cannot help with the task.

This argument is patently false. While it may be true that building only nuclear may not be enough to stave off dangerous climate change, it is clear that nuclear is one of the best single technologies if we are really serious about combating climate change.

To wit, Exhibit A: historical long-term achievements of low-carbon energy, normalised to account for population differences. We used the widely available and generally reliable BP World Energy Outlook data to first search for the largest increase in low-carbon electricity generation over a 15-year window in each of the country covered; we then divided this increase with the average population of the country during the period.

As a result, we can see that nuclear power is surprisingly effective method for increasing low-carbon energy production. All the top spots in this graph belong to 1970s-80s era nuclear programs. Even the much-maligned Olkiluoto 3, the poster child of anti-nuclear movements worldwide, turns out to be faster than any attempt at wind and solar power combined.

Were we to normalise the results according to metrics that are even more relevant – gross domestic product (GDP) or purchasing power adjusted GDP per capita – nuclear power would look even better. The countries shown above succeeded in unprecedented and unequalled increase in low-carbon energy generation back in the day when they were much poorer than today, and significantly poorer than countries with high renewable energy ambitions. As the world’s poor continue to aspire for higher standards of living, this is an important point.

Of course, renewable installations have been growing fast. It is very possible, and very desirable, that some country will one day take the top spot in this chart with wind and solar installations combined. But it is not today; and we’re in a hurry.

One might also want to argue that the 15-year timescale is unfair: the largest increase in renewable energy has happened over the last decade or less. But decarbonisation is not a sprint, it is a marathon. Nuclear plants take time to plan and build, but permits and plans for renewable energy aren’t instantaneous either. We really need to take the long-term view; besides, had we selected the best five-year period (for example), nuclear energy would have won even more clearly.

All in all, we now need all the options at our disposal. At this point, anything else is a huge gamble with our climate.

How to cut CO2 emissions? Four examples (Weekly pic)

Four countries, four histories of CO2 reductions. Data from World Bank and CDIAC carbon database.

In order to avoid dangerous climate change – a calamity that would disproportionately hit the world’s poorest and most vulnerable – we now need to reduce our greenhouse gas emissions at a rate exceeding 3 percent per year for the next 35 years or more. This is a monumental task, to put it mildly.

Given the stakes, the need, and the short time available, one would assume the world leaders would take a look at historically achieved emission reduction rates. However, evidence suggests this hasn’t been happening. How otherwise should we explain that climate discussions still raise Germany’s Energiewende policies to the pedestal, while completely ignoring those countries who have actually managed to reduce emissions at or nearly at the rates now required?

This graph, based on CDIAC carbon emission database and World Bank statistics, tracks the history of per capita emissions and identifies the best 10-year period of emission reductions. As you can see, in this historical light the achievements of that paragon of climate policies seem more like dismal failure than an example to be followed. At least three countries (as well as additional examples such as Switzerland) have accidentally implemented better climate policies than what the “best” climate policy so far has delivered. Two of these, France and Sweden, have produced at least 80 percent of their electricity from low-carbon sources since 1990. This 80 percent happens to be the goal Germany is striving for – by 2050!  What’s more, these policies were enacted against a backdrop of rising electricity consumption per capita, whereas German policy relies on ambitious energy saving schemes coming to fruition.

In every field imaginable, it would be very bad news indeed if accident and chance repeatedly delivers better results than the “best” policy promoted. We need to raise our voices to ensure policy-makers stop burying their heads in sand and take a hard look at the measures that have actually and repeatedly been able to reduce emissions at rates now required. Anything else is a huge gamble with the climate and our living world.

For more information, get our book Climate Gamble, either from online stores or by coming to Paris during COP21 negotiations where we’re handing out nearly 5000 copies for free.

Land use of equivalent wind and nuclear power generation (Weekly pic)

Land use of equivalent wind and nuclear power generation. Examples are Olkiluoto nuclear power plant (Finland) and Ranger uranium mine (Australia), and a composite graphic based on Oosinselkä wind farm (3 MW turbines).
Land use of equivalent wind and nuclear power generation (ca. 27 terawatt hours per year). Examples are Olkiluoto nuclear power plant and Onkalo waste repository (Finland) with Ranger uranium mine (Australia), and a composite graphic based on Oosinselkä wind farm (3 MW turbines). Maps are to scale and include only essential roads and power connections.

Besides climate change, one of the major environmental problems of our age is biodiversity loss. As it is caused largely by increased human land use, stopping biodiversity loss requires us to reduce the land footprint of humanity.

A major problem with renewable-heavy or renewable-only energy scenarios is that they essentially do not acknowledge any problem with human land use. In these visions, vast areas are to be dotted with wind turbines, filled with solar panels or (perhaps most problematically) used for biofuel production to produce enough energy and to deal with intermittency of stochastic energy sources: if you’ve ever heard the explanation “it’s always windy somewhere,” you’ve heard a call to build two to five times as many wind turbines as their nominal production would suggest.

Compared to wind, solar and biomass farms of the required scale, the environmental footprint of nuclear energy is very small. The graph above attempts to show this difference, based on real-world projects in Finland (and an Australian uranium mine). Note that this is for similar annual production of about 27 terawatt hours: to produce energy of similar quality – that is, non-intermittent – one would probably require at least three wind farms of equal size, spaced far enough apart, or significant energy storage facilities.

Note also that the figure likely overestimates nuclear land use. The Olkiluoto plant contains a reserved area for fourth reactor, capable of increasing the annual production from same land area by as much as 13 TWh; and the Ranger uranium mine could easily supply many more reactors.

We would like to emphasize that this picture should not be construed as an argument against wind power or other renewables. We almost certainly need all the low-carbon energy we can have, and both wind and solar have, on the whole, much lower environmental footprint than fossil fuels. The only reason we publish this image is to show that the oft-stated claim of nuclear energy’s environmental destructiveness is misleading at best, and outright falsehood at worst.

This series of posts introduces graphics from our book Climate Gamble: Is Anti-Nuclear Activism Endangering Our Future? The book is now available on Amazon.com in Kindle and paperback formats; see also our crowdfunding initiative which aims to deliver a copy of the book to COP21 climate delegates in Paris this December.

Are renewable installations stalling? (Weekly pic)

How soon are renewables peaking?
Renewable energy installations (nameplate capacity) have recently even declined, long before the build rates required for decarbonization have been achieved. Particularly worrying is the sharp decline in solar PV installations in Europe. Sources: EPIA & GWEC.

The previous weekly pic introduced the calculations of Loftus et al. (2015), which show that decarbonization scenarios that do not allow nuclear energy require stunning, unprecedented rates of new clean energy installations. Even though the popular press is today awash with news of renewable energy achievements, these required rates are still far away. More ominously, there are some indications that the rate of increase in renewable energy installations may be slowing down, perhaps even stalling.

The most prominent example comes from solar PV installations in Europe. Compared to peak in 2011, new solar PV generation capacity is being installed far slower. Subsidies have dried up, and although installations still continue, the major problem is that the rate is far from what’s required for decarbonizing the economy. Furthermore, as solar panels (and other energy generators) inevitably age and need to be replaced, the rate of new capacity addition soon needs to increase even further, simply to replace retiring generation.

It is more than likely that the installation rates will increase from the lows presented here. Nevertheless, one needs to remember the previous post’s message: if we want to decarbonize without nuclear power, we need absolutely huge increases from current installation rates. It bodes ill for the prospects of these rates being achieved that these hiccups occur already, when solar and wind together still provide less energy to the world than nuclear power alone.

Nevertheless, some members of our society still think the required increases in renewable installations and energy savings rates are done deal, and that we can forget about nuclear power entirely. These graphs point out again that this stance is a huge gamble with the climate.

This series of posts introduces graphics from our book Climate Gamble: Is Anti-Nuclear Activism Endangering Our Future? The book is now available on Amazon.com in Kindle and paperback formats; see also our crowdfunding initiative which aims to deliver a copy of the book to COP21 climate delegates in Paris this December.