Burden of proof

Ted Trainer, with whom I’ve had a number of debates in the past, has sent me an interesting piece claiming that “no empirical or historical evidence that demonstrates that [100 per cent renewables” systems are in fact feasible”. The authors, at least those of whom I’m aware, are “pro-nuclear environmentalists” (Ben Heard, Barry Brook, Tom Wigley and CJ Bradshaw) The central premise is that, given that renewables won’t work, and reductions in energy demand are unrealistic, we need to get cracking on nuclear (and also carbon capture and sequestration).

It’s paywalled, but the abstract is sufficient to get the main point. In fact, the whole piece is summarized by its title “Burden of Proof”. To give the shorter version: Unless every possible detail of a 100 per cent renewable system can be proved to be workable decades in advance, we must go nuclear.

The longer version is in these paras from the abstract

Strong empirical evidence of feasibility must be demonstrated for any study that attempts to construct or model a low-carbon energy future based on any combination of low-carbon technology.

The criteria are: (1) consistency with mainstream energy-demand forecasts; (2) simulating supply to meet demand reliably at hourly, half-hourly, and five-minute timescales, with resilience to extreme climate events; (3) identifying necessary transmission and distribution requirements; and (4) maintaining the provision of essential ancillary services.

This list is mostly notable for what’s not in it: adequate year-round power supplies, at an economically feasible cost. That’s because it’s now obvious that solar PV and wind, combined with one of a number of storage technologies (solar thermal, batteries, pumped hydro) and a bit of smart pricing, can deliver these goals. So, instead we get demands for the precise details in the list above. To lift the burden of proof a bit more, it’s not good enough to address them separately, they all have to be done at once in a single study. Unsurprisingly, no one has yet produced a study that meets all of these demands at once.*

And this is where the burden of proof works so brilliantly. Renewable technologies are well established, with annual installations of 100 GW a year a more, and a record of steadily falling costs. But, according to our authors, they haven’t met the burden of proof, so we have to put tens of billions of dollars into technologies that are either purely conceptual (Gen IV nuclear) or hopelessly uneconomic on the basis of current experience (CCS and generation II/III nuclear).

To be fair, this use of the burden of proof, while more blatant than usual, is very common. One any policy issue, most of us would like to compare an idealised model of our preferred solution with the worst case scenario (or, at best, the messy and unsatisfactory reality) for the alternatives. But it’s important to avoid this temptation as much as possible. On any realistic assessment, renewables + storage (with the path to 100 per cent smoothed by gas) offer a far more plausible way of decarbonizing electricity generation than nuclear or CCS>

Clarification: In comments, Ben Heard points out that the authors counted two publications from closely related studies together.

81 thoughts on “Burden of proof

  1. @James Wimberley
    IMO the key thing CSP (CST+storage) has to demonstrate for any proposed location — particularly if not wedded to sufficient pumped hydro — is that it has the capability to ride through all of June July August (southern winter), given the known worst case of protracted overcast weather in the historical meteo records. *AND* (during average winter conditions) be able to meet a guaranteed capacity factor, say 200MW continuous for 260MW of installed capacity. That might be 200*24*3 for three days winter energy export (14.4GWh), or a molten salt energy store of 36 GWh in the form of heat (40% efficient heat-to-electricity). ….that surely must be a very big tank of molten salt to supply 130 tera-Joules (1GWh=3.6TJ i think).

    The beauty of “wind farms fully integrated with pumped hydro” is a better ability to ride through a typical winter in SE Australia: the wind regime is typically low in autumn but picks up in winter, even moreso for offshore wind farms, with their much better night-time energy export capability. Offshore wind never suffers from a temperature inversion at night, so the surface winds never decouple from the weather systems in the mid-tropospheric levels. Steady nighttime winds mean Australia can decarbonise its power grid much faster with offshore wind than with the same installed capacity of onshore wind. With offshore wind giving better night-time capacity factor in all seasons, the TWh of NEM’s aggregate pumped hydro “virtual battery” can be considerably downsized.

  2. @derrida derider
    It has been known by experts for at least a decade that w/o any oil or coal combustion, the known reserves of PETROMETHANE (cutely spun as “natural gas”) were enough alone to push Earth over 2degC into runaway climate change, when they are all burned. Leggett (see YouTube clip below) may have allowed a small fugitive emissions scenario, but he may also be saying that w/o any factoring in of unburned CH4 escaping we’re all gonna fry: https://www.youtube.com/watch?v=dM656m8TeyY

  3. @John Quiggin
    Not a matter of relaxing anything. Again, if you read the paper closely, you can see that where authors sought to respond to a criteria, or describe a matter as having been addressed and note to be covered in the paper, we reviewed those sources. In the example we discuss in the paper, this was entirely unsatisfactory as the reference in no way adequately supported the further work.

    The study is not directed principally as a criticism of the studies we reviewed. In some cases that criticism may be applicable, and in others (Elliston et al as we highlighted) is certainly isn’t even though the score was low. The paper is directed at the notion that literature affirms that 100% renewable electricity systems are feasible with the necessary confidence that policy makers should or could factor out other forms of low carbon energy.

  4. I think Michael Gunter’s point is excellent. It is not yet clear which will be the most cost-effective way to deal with the intermittency of solar and wind over the southern winter.
    And the other complication is electric transportation. We will need twice as much electricity generated as we generate now to electrify transport so that is a challenge, but the batteries in electric transport will give us a means to deal with intermittency as long as we develop the appropriate smart systems and economic incentives.
    So certainly there need to be some clever solutions worked out to ensure that a 100% renewable energy system will not be excessively expensive, but there is no question as to whether we can achieve a 100% renewable energy system.

  5. @Ikonoclast

    Also, the “irregular angular shaped areas” in between circular fields are not the show-stoppers you seem to imagine them to be.

    1. The final formula for the answer = r squared x (4 – pi). Disclosure, I looked it up rather than dust off my rusty Senior maths.

    I’m unsure which circles packing problem that one is for. Some number bounded by a square? The following two formulas ought better suit you. The highest efficiency (also looked up) apparently is the hexagonal lattice arrangement on an _unbounded_ plane: ηh = π/2√3 ≈ 0.9069 (wiki); ηh = 1/6π√3 ≈ 0.9068996821 (wolfram)

    But that is on planet maths. On this planet ‘ought is not is’. Why is that so? Numerous factors… some relevant to farming were mentioned earlier. Look again at your satellite image.

    2. This area is not “dead” or unusable. Uses
    The uses mentioned are not aligned with industrial scale modern agricultural practice which turns oil into food and currently can feed the world – more or less.

    It is not the case that everything we do next has to have been already done. We can innovate after all.

    “Build it and they will come” – a common misquote from the movie Field of Dreams – a fundamental observation of the transition engineering needed NOW to avoid chaos and collapse – see Susan Krumdieck. People generally expect of tomorrow what they expected of today, their expectations change because of change built about them.

  6. @Michael Gunter
    OK, so if CSP is your main supply, you have 3 days a year when you have to turn on a backup gas generator. (I assume they get overcast days in Nevada, if not the rainless Atacama). 99% renewable is good enough for government work. It’s better anyway than the availability of nuclear, where shutdowns for maintenance are unavoidable. In practice, security of supply will be met from diversity, and the degree of security will as now be balanced with cost.

  7. … increasing in scarcity.

    God christ did you read a single thing I wrote? Did you understand it, or are you just running some sort of bullshit spinal reflex in your responses here?


    Agricultural land is a manufactured commodity. It is made from:
    + dirt
    + water
    + nutrient chemicals of various sorts
    + chemicals to ameliorate negative elements of the soil [acidity regulators, etc]
    + things to improve the structure and physical properties of the soil so that the desired crop can grow better [mulch, tilling, etc]

    Of these things, growing things in circles with gaps between them is wasteful, but it’s only wasteful of the first thing, the physical space. About ten percent in the best case, more plausibly fifteen [but as has been pointed out the “wasted” land can be used for other purposes; the actual loss is less than that].

    But — and this was the reason I made my post — things other than land aren’t wasted that way. We only till and fertilise the land we’re actually using, not the gaps between; there’s no increase in costs other than costs-of-physical-dirt.

    And we aren’t short of dirt. We have a planet full of it. Much of it isn’t currently ready for intensive cropping… but land-ready-for-cropping is something we make, out of normal land, of which we have no shortage as mentioned.

    A fifteen percent increase in land-used-for-intensive-cropping… this isn’t actually an insurmountable problem, because physical dirt isn’t anywhere close to the limiting resource on agriculture.

    [thus: centre-pivot irrigation: the land between isn’t wasted; before centre-pivot irrigation it couldn’t be used for cropping at all. It’s all bonus.]

    Land suitable for farming is — at present and for the foreseable — not something that’s inescapably “increasing in scarcity”. We have as much of it as we’re willing to make, and a fifteen percent increase in physical space requirements for cropping is not a vast problem.

    [but all this is extra-special stupid because we can just use longer power cords; there’s no requirement that areas accessible by tractor be non-overlapping!! Fifteen percent increase in the length of the cord — copper cable is not very expensive — turns your waste land into doubly-accesssible land and makes your problem go away.]

    [I have spent twenty minutes telling you things you shouldn’t need to be told. At my boss’s charge-out rates for me that’s… about sixteen dollars of cost you’ve put the planet to. Show me you’re worth it.]

  8. Svante,

    The solution I gave is simply for a circle in a square. Thus, I implicitly accepted squares would lined up in grid fashion. You note that a hexagonal “meta-field” layout is more efficient for the circles with less space wasted for the prime purpose. It makes sense and so much the better.


    I would not so rapidly assume as you do that there is no use for the “dead areas” (my term). There will almost certainly be uses consistent with industrial scale modern agricultural practice. I mean both now and as it could develop in the future.

    Have you ploughed and sown large, rectangular dryland farming wheat fields? I mean fields in the size vicinity of 600 to 1,200 acres at least. I have in the Shire of Mingenew, Western Australia, about 4 decades ago. I wonder, do you imagine this form of farming was/is all completely efficient? (Yes, I know I needed a very large diesel tractor with doubled-wheels on the back. I will come back to that.) For a host of reasons, this form of farming has its own inefficiencies. Among them is the issue of ploughing and cultivating the “headlands”. I will let you look up Plowing Headlands on Google or Youtube. Woe betide the rookie who turns the wrong way. I did… once.

    When one gets right down to it, a circular field (of ample size on a flat enough or leveled plain) would bring all kinds of efficiencies. With modern technology, a boomed, tethered or even free, self-driving rig (an Unmanned Agricultural Vehicle) could very efficiently plough or cultivate a circular field. No need for a driver. The operator would probably set up the run and then manage “pit-stops” for seed grain and/or fertilizer when cultivating… oh hang on, these could be delivered by a closed pipe (air-propelled) or closed conveyor on an extending boom system.

    You say “People generally expect of tomorrow what they expected of today.” I mean really m8, who does that describe… me or you?

  9. Or you could make the power cords 15% longer, and let the circles overlap.

    [or you could realise that physical-space is only a part of the cost of agricultural land, and a drop in land uitilisation of 10% doesn’t thereby mean a drop of 10% production for the same inputs. Or column A / column B, or what-have-you.]

    I mean, the reasons the circles with centre-pivot irrigation don’t overlap isn’t because it’s technically impossible, or even particularly difficult, it’s because land is so cheap that it’s not worth the hassle: the ongoing maintenance costs for the synchronisation gear and water-flow controls, etc, are more than the value of the extra production you could get from the “wasted” land in between.

    Oddly, I’m reminded that the same thing applies to politics: some people cost so much to explain things to that you’re better off leaving them in ignorance.

  10. Less oil, no till; no alt-oil, no way. Behold, crop circles, the way. Where flat earthers are wont to plough up cloddy shibboleths all over round and round again there’ll be no harrowing after.

    Same for nukes.

  11. @Mark Duffett (@MarkDuffett) The burden of proof is legitimately lighter for nuclear because it doesn’t rely on models, it can point to reality

    So Mark, which countries use 100% nuclear electricity? That’s the standard Ben requires and you seem to be supporting.

    And as I pointed out above, Te Waka o Maui regularly gets 90%+ (average 98%) renewable electricity, but since it’s connected to the rest of the country and it’s cheaper to export in the winter and import (partially) fossil electricity in the summer they do that. It’s not because they *can’t* run 100% renewable, but it’s more profitable not to. Tweak the subsidies the other way and they’d like be running 120% or more renewable (limited by transmission capacity).

    I think it’s worth noting that we don’t have to switch abruptly to 100% anything for the electricity supply, and it would be expensive to try. What we can and should do is move incrementally away from polluting power sources, keeping whatever requirements from the current system make sense. That’s how other countries have transitioned between power sources.

  12. @John Goss
    John Goss perceptively asks who reviewed Heard et al. (2017). At a recent international conference on ‘Energy for Society’, Michael Jefferson, an openly pro-nuclear climate sceptic and renewable energy sceptic who co-edits the international journal Energy Policy published by Elsevier, revealed in a Q&A that one of his energy experts (reviewers?) is Ted Trainer, a retired social work academic. Was Ted was one of the reviewers of the Heard et al. article, which was published in another Elsevier journal, Renewable & Sustainable Energy Reviews? BTW, I used to publish frequently in Energy Policy, but will no longer submit papers there while Jefferson is editor.

  13. @Moz of Yarramulla

    Who is asking for 100% nuclear? To be clear, the argument is over why people insist on there being 0% nuclear. And is not even the important question; which would be how do we best get low gCO₂/kWh? If you have access to excessive amounts of hydro (& geothermal) then you probably should be up around 100% renewables, and for transport too. People can argue till the cows come home over research papers & the methodology under which they are assessed (though I haven’t seen a lot of objection to the actual criteria used in the Heard el al paper)… and then retreat to their previously held convictions.

    Meanwhile, over in Ontario: 7gCO₂/kWh (https://twitter.com/ParetoEnergy/status/851669758382452736) Which would solve a whole bunch of problems if repeated around the world.

  14. I don’t participate in peer review enough to know for sure, but it seems to me from observing the outputs that the peer review system is broken. Because the university rankings depend mostly on papers published/ cited, academics are submitting as many papers as they can, so the load on peer reviewers is enormous, so that would be bringing down the quality of peer review. And if a paper is rejected by one journal then one keeps submitting it until it is eventually accepted, and this again adds to the workload for peer reviewers.
    And there is less incentive for academics to do peer review as you don’t get much credit for it in performance reviews, as number of peer reviews don’t get into the rankings.
    So its a vicious cycle, and again inappropriate performance incentives seem to be to blame.

  15. @John Goss
    I was involved in a paper submitted to journal A, which was rejected outright by the editor. The paper was submitted to journal B, whose editor thought it was good work but not quite suited to the journal. The editor suggested journal A and sent journal A’s editor a note recommending that the paper be considered. On second submission to journal A, the paper was accepted with no changes.

  16. Have the nuculistas explained where the bottomless pit of money for their pet hobby is located because I’d like some too?

  17. Don’t nuclear facilities have major insurance problems? Doesn’t the State always end up having to pass legislation that shifts risk from the nuclear cowboys to Joe and Jane Public?

  18. @Mark Duffett (@MarkDuffett)
    “The Costa Rican Electricity Institute (ICE) said that around 98.1 per cent of the country’s electricity came [inn 2016] from green sources.” Source.

    I am using the commonsense definition of “renewable” as including hydroelectricity. The weather system replenishes the water in the rivers without human intervention. That does not mean that hydro does not have its environmental problems, see Belo Monte and Narmada. Sure, it’s not available everywhere. Nor is anything else.

  19. @Jonathan Sutanto
    Thanks for pointing out that nuclear has fared quite well in the CO2 reduction strategy. I’ll also point out that subsequent Renewables Procurements Plans undertaken as a result of the Green Energy Act are now admitted to be a mistake by the same government that introduced the GEA and the resultant RPP waste of money, time, and resources to Ontario. The article mentions nary a word as to the reason Ontario hasn’t sunk further: clean reliable 60% nuclear.

  20. I’ve just thought of another point of framing bias. Why pick 100% renewable? It’s a political aspiration and policy goal, sure, so worth thinking about. But it’s not an operational problem for a long time. The decisions taken now are about moving the electricity supply system to much higher levels of renewable penetration than we have today. For the sake of argument, let’s say 90%, though no large country reaches this at present. (Brazil is at 82%, mostly hydro). It’s obvious that allowing this amount of slack greatly widens your options, as you can use gas peakers. On general grounds, the last 10%, 5% and 1% of decarbonisation will be much more difficult and increasingly expensive. It may well be more cost-effective in emissions terms to put the money into something else like electric trucks or forestry. In any case, it is quite reasonable to leave the last-mile problem till later, when the technical options will have moved on. Just picking 100% renewable and finding arguments suggesting it’s difficult is a straw man argument.

  21. @James Wimberley

    You are being logical again! 🙂 You know how the nuclear shills and boosters dislike logic. How can you be so unfair as to be logical?

    Seriously, of course you are correct. 100% anything is pretty much impossible in human endeavours. Try making anything 100% pure (not one part per billion billion of any contaminant). Try making anything absolutely 100% safe and so on. It’s an absurd metric only set when the PTB (powers that be) wish to block something they don’t want to happen. Sometimes they block for reasons of vested financial interest and sometimes they block for ideological reasons (which really comes down to retaining power).

  22. @James Wimberley

    While I take your broad point — that ‘the last mile’ can be deferred, I’m not sure it will generally prove to be the most expensive or technically challenging. My own perception is that the first 50% of decarbonisation has been far more difficult. All the technologies and their interfaces with the grid had to be reconciled with a system designed to maximise industrial, commercial and residential demand at timesof convenience to ‘baseload’ plants.

    All of the new technology is/was in a process of refinement as the whole energy market is/was changing. Yet both storage and energy harvesting measures are improving sharply each year. Recently, bids on offshore wind in the North Sea fell to around €4.4MW for wind. At least one offered to supply at €0 — taking their chances that the wholesale market would bail them out.

    Really, I’d sooner limit FHC use to settings where it really is unavoidable even if doing so came initially at a premium price. There are ways in which CO2e can be drawn down and sequestered long term — e.g. algae … But we are about 1500GT of atmospheric CO2 above what is in the longterm interests of humanity, and seemingly about to lose large chunks of our glacial mass and permafrost. We must not allow that to happen, or treat mitigation as an e istential project on the only place in the universe where we know life exists. Every cost that forecloses this probably unique ecosystem from collapse is warranted — and would be even were a consequence that our quality of life would decline in some measurable way.

    Since nuclear power has been raised, some will recall that, like Derrida above, I regard the objections on safety as exaggerated. The other objections – principally schedule and mission (and in Australia, technical and regulatory) feasibility — seem to me to be far greater problems than safety, or even cost.

    While I don’t favour forced decommission of existing nuclear plants that are fully compliant with all reasonable safety provisions (particularly in the light of Fukushima) and regard bulk container shipping as a plausible target for nuclear engines — these ships are a terrible source of both CO2 and other marine hazmat — it seems very clear that if nuclear had a window of opportunity to grow, it was missed — probably in the late 1970s. By 2012 other low emission technologies had clearly elbowed their way in front and now continue to put nuclear out of any contest for market share of the domestic stationary energy market that could conceivably arise.

  23. @James Wimberley
    Detractors have previously said that if we cannot get to 100% renewables, we need to choose a different path right now as we are barking up the wrong tree. The point of 100% RE scenarios is to test and evaluate the limiting case. Call it a framing bias if you like, or call it pushing the envelope.

  24. I don’t know if this paper is intended to invoke discussion, or simply advocate nuclear generation. Had the world stridden that path years ago we might still be needing solar/wind, but less urgently. That would have been nice but we can’t go back.

    Two thoughts spin in my head: necessity is the mother of invention; politicians don’t see the necessity.

    What I would find helpful is something like a sequel to the book by the late David MacKay FRS:

  25. James Wimberley :
    Why pick 100% renewable?

    Because it makes the task of the “opponents” much harder. In theory we’re on the same side, but in practice I think there’s a bit of “when all you have is a hammer” going on.

    As you saw above when I suggested that the obvious comparison was with countries that have gone 100% nuclear … the comment was laughed down. Which is a suitable response, IMO, to any of the 100% demands. 100% requires insanity at the margins, for example collecting and disposing of carbon dioxide and methane emissions from freshly felled biofuel plants, not to mention the need to completely automate everything in order to eliminate emissions from humans.

    The amusing thing is that it’s much easier to get to 100% renewable than any other broad field except “zero emissions” since that’s a superset.

  26. Now 410.28ppm C02. (I though you said the age of increase – and increasing increasinbng – was over…)

  27. @Dennis Horne

    I read the Mackay text some years ago, and was on the whole, impressed with it. It does make some plausible claims for nuclear power.

    The problem with the text is that its focus is, on the whole, concerned with the engineering or technical feasibility questions — which is scarcely surprising, given Mackay’s background.

    What it fails to deal with, and which is now considerably more plain than it was when Mackay composed it — were the falling cost of renewables, the speed with which RE can be rolled out, relative to nuclear, and the associated risk premiums on bankrolling new nuclear power given the uncertainties if both the approval processes and the longterm markets for power over the likely working lives of plants.

    In effect, bankrolling plants entails taking the longest of long positions over a very slippery commodity — power supply. Only states can contemplate that, which is why to keep nuclear vaguely competitive in a market where energy production costs are sliding sharply, those places where nuclear development is taking place are either wholly owned by states or given feed in guarantees. Yet in theory, most states avow ‘free market principles’ and in the UK and the US much of the ruling class is against an explicit price on carbon emissions.

    More generally of course, solutions to a low carbon power supply are needed far earlier than in the period 2032 to 2092 — the earliest that most nuclear projects in western countries could begin supply, so Mackay’s argument is rendered entirely moot by the failure of the technology to meet schedule and political feasibility.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s