David MacKay: what energy portfolio would he favour today?


The UK is making more smart energy policy than most of the rich world (who do you think is #1 on the rankings of effective climate change policy?). One of the most remarkable smart things the UK government did was to appoint prof. David MacKay as Chief Scientific Advisor to the Department of Energy and Climate Change (DECC) for the period 2010 to 2014. I consider Dr. MacKay one of the top energy policy authorities on planet Earth. It’s unusual when a government selects someone so qualified to such an important position. Even more unusual when the selectee is respected for speaking his views regardless of political consequences.

One of the important deliverables that David MacKay initiated is the Global Calculator where you Dear Reader can experiment with your preferred energy policy options — to see what the consequences are likely to be, what “adds up”. And if you don’t like the model outputs, then you can download the complete code as an Excel spreadsheet. So you can produce your own more-perfect model of how the planet, and planetary economy will respond to your energy policy.

Back to the captioned question “David MacKay: what energy portfolio would he favour today?” I don’t know what Dr. MacKay would propose if he were invited to design the UK (or Earth) energy policy. But I’m completely confident that I would prefer his policy to anything likely to emerge from the UN process.

Today a few clues of the David MacKay preferred policy surfaced on Twitter. I’ll summarize what caught my attention – the conversation I captured is here. I’ll summarize:

Bishop Hill @aDissentient asks “Is there a rational explanation for why governments keep pursuing wind power? One assumes it’s *despite* your book and advice”

David MacKay (@davidjcmackay) replies [in 5 tweets] One rational explanation is “as an option”. I think a hedging strategy is wise given how difficult all the levers are. Another explanation is “because of the legally-binding EU renewable target that Tony Blair signed UK up to“. I concur that I think best vfm hedging strategy wd now steer the grid mix more strongly to non-intermittent low-carbon. [ie CCS/N/very-large-tidal] – though this would lead to legal infringement of EU RT (see pt 2). If there is a huge R+D breakthrough on storage, however, the optimal hedging grid-mix might change.

The most obvious flaws in the OECD policy generally are the special economic/regulatory incentives granted to the “soft path” options of solar/wind/biofuels. I read Dr. MacKay’s “best vfm hedging strategy wd now steer the grid mix more strongly to non-intermittent low-carbon” as an explicit recognition of the real system costs of intermittent generation. And the need to pay more attention to the realities of reliable power, and less attention to “feel good policies”.

PS: When you think about UK energy policy were you aware of “because of the legally-binding EU renewable target that Tony Blair signed UK up to”? I was not. Be careful what you sign!

I hope you will explore the Global Calculator. It is a remarkable resource. Thank you UK Taxpayer.

9 thoughts on “David MacKay: what energy portfolio would he favour today?

  1. The Global Calculator in your link has a pro-renewables-anti-nuclear ideology baked into its assumptions.

    The Calculator assumes that the maximum rate at which nuclear generating capacity could be built in a global all-out effort over the period 2012-2050 would be ~38.5 Gw. Under this assumption, they claim that over that period, we could add about 1.5TW of capacity, which they show would amount to a small contribution in solving our GHG problem.

    This is an odd assumption for them to make, since a build rate of 38.5GW/yr isn’t much higher than the 30GW/yr build rate achieved during the 1980s. Most of that build out took place in the US and France, and was the result of an effort that could hardly be characterized as global or all-out. Barry Brook, at Brave New Climate, estimates that if the level of effort that went into the 1980s roll out of nuclear were duplicated today, and on a truly global scale, it would result in a deployment rate of ~170GW/y, almost 4.5 times faster than the Global Calculator team assume.

    By comparison, the Global Calculator team assume that solar capacity, if deployed in a global all-out effort from 2012-2050, could be added at an average rate of ~590GW/yr. This would represent something like an order of magnitude increase in the maximum historically achieved rate of deployment. The technical challenges of integrating a variable supply input on this scale appear to have been essentially assumed away. Under their set of unstated assumptions, Gobal Calculator “shows” that solar, more than any other energy technology, is capable of playing a major role in limiting GHG-induced warming to 2°C. Again, this is odd given that, of all the “clean” energy technologies available, only geothermal has made a contribution smaller than solar.

    I’ve noticed that Greens rely heavily on arguments in which the conclusion is assumed. It looks like Global Calculator might be another example of this sort of tactic.

  2. I get different numbers but similar conclusions to yours using this source Global Calculator Sector metrics from 2°C pathways. Here are my numbers – extracted from page 4 “Electricity generation and fuels”:


    Barry Brook’s ~170 GW/y is entirely reasonable as an average build rate (to 2060 in his model). If my calculations are correct, GC is projecting a maximum average annual nuclear build rate of 17 GW/y. That’s a full 10x slower deployment rate than Barry’s.

    By comparison solar and wind are both projected to average build rates around 6 to 7 times faster than nuclear. That makes no sense on a technical basis. It could be a completely accurate projection of the political results.

    Projecting how politics will play out over 4 decades is a hazardous game. But my guess is that 70% plus of the new generation built in this period is outside OECD “feel good” politics. China is going to build what works in the field – and it will not be high-penetration wind/solar.

    • Hi Steve,

      I got the numbers I cited straight from the calculator and the notes that accompany it, not from the spreadsheets of the underlying data.

      The numbers you’re getting from their scenario for nuclear come from their “level 3” projection, which they characterize as “very ambitious”. In their level 4 (“extremely ambitious”) projection, world nuclear capacity rises from 364GW of capacity in 2011 to 1870GW in 2050, which implies an average deployment rate of 38.62GW/yr. In the notes, GC describes this scenario for nuclear like this: “After 2020, the build rate increases to the fastest the nuclear sector believes is technically possible, and this is sustained until 2050. This would result in 1,870 GW of capacity. This represents a future which has put an extreme amount of effort into rolling out nuclear.” (http://tool.globalcalculator.org/gc-lever-description-v23.html?id=26/en)

      Their figures for solar that you see in the calculator and in the supporting notes differ. I used the figures they show in the calculator, since that’s what will be doing the most to inform public impressions, and since those projections also appear in their spreadsheet. In their level 4 scenario, solar capacity rises from 70GW in 2011 to 23,107GW in 2050, an implied an average deployment rate of 590.69GW/yr. In the notes (where they instead state a final solar capacity of 4817GW by 2050) they say: “Both solar PV and concentrated solar power have improved and new technologies, such as printable photovoltaics, may be a big part of the market. Storage capacity is also increased significantly. Level 4 is based on Shell’s “Oceans” scenario.”

      I don’t see any of their deeper underlying assumptions about solar or nuclear stated explicitly anywhere, so it’s hard to comment on any specific aspect of them. Generally though, they seem to allow for hypothetical solar, in which various major technological breakthroughs and resulting technological capabilities are assumed. Similar assumptions seem to have been made for wind.

      On the nuclear side, judging from what they assume about the peak rate at which a concerted global effort would be able to deploy nuclear capacity, I think it’s safe to say that they’re assuming future nuclear technology, production and construction methods will not have advanced beyond 1970s or 1980s levels by 2050. My guess is that they’re also assuming restrictive regulatory environments around the developed world remain in place, in spite of the fact that Global Calculator purports to model what could potentially be achieved if humanity proved capable of taking decisive actions to combat global warming.

      For my part, I think the level 4 scenario for nuclear depicted in GC would more accurately be designated a level 2 (although GC are a bit vague about what distinguishes the various levels from one another). All that would be required to achieve a result very close to what is portrayed in their level 4 nuclear scenario would be for the following nations/regions to deploy the equivalent of one 2GW NPP every year over the period 2019-2050: USA, Canada, Central America, South America, UK, France, Germany, Italy, Iberia, Scandinavia, Rest of Western Europe, Russia, Rest of Eastern Europe/Former Eastern Bloc, Africa, Middle East, India, China, South Korea, Japan, Rest of Asia, and Oceania.

      This scenario, which would be something like a concerted but minimal global effort, would result in the addition of 1344GW of capacity added at an average rate over the period 2012-2050 of ~34.5GW/yr (assume that a decision to pursue this strategy is taken in 2012, and the first plants come online 7 years later). Historical evidence strongly indicates most of these nations/regions could achieve deployment rates several times greater than the one I’ve just outlined without placing any undue strain on their respective economies.

      Again, given already demonstrated nuclear capacity deployment rates in the 1980s of 30GW/yr, it’s hard to understand why anyone would assume in good faith that a global effort of similar intensity, if undertaken today, should have any difficulty exceeding this result by a wide margin. Of course, as others have pointed out, good faith is often in short supply. (https://passiiviidentiteetti.wordpress.com/2015/03/03/short-list-of-dishonest-and-silly-arguments-in-energy-discussions/)

  3. The Global Calculator team replied to my inquiry on Twitter. Here’s the conversation up to the following reply:

    “@GlobalCalc: @stevedarden @ActinideAge @davidjcmackay I’ve checked, and experts agreed nuclear L4 based on reports from Oxford and MIT.” 

    I hope Oxford doesn’t refer to the Oxford Research Group who were referenced in David MacKay’s Sustainable Energy: Without the Hot Air (pg 171):

    “For nuclear power to make any significant contribution to a reduction in global carbon emissions in the next two generations, the industry would have to construct nearly 3000 new reactors – or about one a week for 60 years. A civil nuclear construction and supply programme on this scale is a pipe dream, and completely unfeasible. The highest historic rate is 3.4 new reactors a year.”

    • I found the MIT paper I’m pretty sure they said they cited. You can find it and its update here: http://web.mit.edu/nuclearpower/pdf/nuclearpower-full.pdf, http://web.mit.edu/nuclearpower/pdf/nuclearpower-update2009.pdf

      The figures 1000GW and 1500GW of nuclear capacity by 2050 do appear in the paper (and they reflect total capacity by that year, not total capacity added). But those figures do not represent what MIT thought to be the maximum conceivable quantity of nuclear capacity that could be built by that time, as GC claim. In fact, the MIT paper never attempts to project how much nuclear capacity could be built, because that wasn’t its purpose.

      The purpose of the 2003 paper they cited was to analyze, “what would be required to retain nuclear power as a significant option for reducing greenhouse gas emissions and meeting growing needs for electricity supply.” (Executive summary, pg ix) In their analysis, they focus on cost, waste, safety and proliferation, four areas they consider to be key challenges that prevent nuclear power from playing a larger role in supplying humanity with energy.

      In order to quantify the four challenges they wanted to look at, they assumed 2050 nuclear capacities of 1000 to 1500GW (that’s total capacity, including the 366GW that already existed at the time of writing). They didn’t choose these values because they thought they reflected upper limit of what is technically feasible, a subject the paper never broaches. Rather, in their words, “We adopt 1000 to 1500 GWe as the mid-century reference point range for our study. This is large enough to reveal the challenges that need to be faced to enable the large-scale deployment of nuclear energy.” (pg 25)

      The closest this paper ever comes to offering an opinion about the rate at which nuclear capacity could be built comes a little further down the page, where they note, given their assumption that global electric power demand will increase slowly at first, and then extremely rapidly by mid-century, in order to keep up with demand and thereby approximately maintain its current share of electricity production, nuclear would need to be built at a faster rate than what was achieved during the 1980s. They feel this built rate would be “challenging”. They don’t elaborate on what they thought “challenging” meant.

      If this is what GQ are using as the basis for their claim that 1500MW represents the upper limit to what “the nuclear sector believes is technically possible”, then they’re making a huge leap. And they’re consulting MIT, not the nuclear sector.

      If they wanted an opinion from the nuclear industry, one place they could have looked would have been the World Nuclear Association’s website. There, they say “It is noteworthy that in the 1980s, 218 power reactors started up, an average of one every 17 days… The average power was 923.5 MWe. So it is not hard to imagine a similar number being commissioned in a decade after about 2015. But with China and India getting up to speed with nuclear energy and a world energy demand double the 1980 level in 2015, a realistic estimate of what is possible might be the equivalent of one 1000 MWe unit worldwide every five days.” Also, “A WNA projection shows at least 1100 GWe of nuclear capacity by 2060, and possibly up to 3500 GWe, compared with 373 GWe today.” A summary of that projection is available here: http://www.world-nuclear.org/uploadedFiles/org/nco/NCO.pdf Sadly, a more detailed version of it seems to be behind a £940 paywall (downloadable here: http://www.world-nuclear.org/WNA/Publications/Publications-for-Sale/Global-Nuclear-Fuel-Market-Report/).

      Using the WNA numbers and using the GC timeframe (so starting in 2010 instead of now), this would mean that by 2020, the world would be adding 73GW of nuclear capacity each year. Assuming that production remains constant for the following 31 years, that would mean 2263GW added over the period 2020-2050, plus whatever is added over the “tooling up” period 2010-2019 and whatever remains by 2050 of existing 2010 capacity.

      As to the “Oxford” paper GC mentioned, I’ve looked, and so far I haven’t found anything from Oxford University that addresses this topic. I found a lot of stuff on issues nuclear from “Oxford Research Group”, and none of it passes the laugh test. I too hope they didn’t cite them.

      • Thanks Andrew, good work. I suppose the MIT Future of Nuclear Power (and 2009 update) is a possibility – though, as you said, the MIT effort was not intended to address an “extremely ambitious” build out. MIT did not seriously examine mass manufacturing of advanced reactors. The MIT report was basically a muddle along on LWR, try a few new designs like HTGR, and negative on closed fuel cycle designs like IFR, Transatomic, etc. The update closed with:

        After five years, no new plants are under construction in the United States and insufficient progress has been made on waste management. The current assistance program put into place by the 2005 EPACT has not yet been effective and needs to be improved. The sober warning is that if more is not done, nuclear power will diminish as a practical and timely option for deployment at a scale that would constitute a material contribution to climate change risk mitigation.

        As I wrote, I hope GC team didn’t use anything from the Oxford Research Group. Given that David MacKay specifically called out their obvious anti-nuclear bias I wouldn’t think they would use that Oxford. 

        The WNA Nuclear Century Outlook Data tabulation by country is as good a mainstream projection as I can think of. It reflects feedback from each country. The tabulation totals are:

        2060 Low 1140 to High 3688

        2100 Low 2062 to High 11046

        I made a little plot to compare the relative growth rates, using an exponential trendline to estimate what the WNA would estimate for 2050 – to align with GC 2050 values.


        Given all the political and technical uncertainties of projecting to 2050, 2060 I’m not sure there’s enough difference here to devote a lot of energy to the nuclear projections. I certainly agree that the VRE projections seem to assume that all sorts of advances will appear when needed, especially storage or some other magical thinking about intermittency. That is in harmony with the current political framework. And no doubt reflects the intense lobbying by all the VRE interest groups. What “interest group” would even call on the GC team to discuss what nuclear can contribute to displace coal?

        There is a lot of value in the Global Calculator model in that it allows citizens and policy people to quickly experiment with a whole range of policy mixes. I’m depressed that the calculator parameters are set to force a total focus on VRE which will certainly lead to a reality-collision around 2030 – 2040.

        Nearly nobody in the policy community is asking “What must we do to ensure that we succeed?”


      • As a hypothetical medium-effort scenario for the USA, a while ago I proposed a factory or set of factories turning out 2 NuScale units (or equivalent) per day, 250 days a year.  That is 95 MW(e)/day * 250 days = 23.8 GW(e)/year.  This would ship about 1400 tons of finished NSS systems per day, a small fraction of the shipments of light-duty vehicles in the USA.

        At ~24 GW/yr, the coal-fired base-load capacity of the USA would be fully replaced within 10 years.  The NG-fired base-load would quickly follow.  Further, a walk-away safe unit like NuScale is suitable for installation inside cities.  The 110 MW of waste heat per unit could displace over 8 million scf/d of natural gas in uses like space heating and DHW.  Multiply by 250 units per year, and each year’s production could knock off 2 billion cf/d, around 700 billion cf/yr.

        Total US natural gas consumption is around 26 trillion cf/yr.  Chewing into that total at a rate of several percent per year, while eliminating coal, would go most of the way to meeting our 2050 climate goals.  It would handily beat the 15%/decade we need, and the effort required (for the NSS side) appears to be on the easy/moderate end.

        I’m not sure what the infrastructure end would require (big installations of steam piping in cities, likely massive reconstruction of most utilities and streets in the process) but at the very least you’d get high employment out of it.

Comments are closed.