Friday, January 26, 2018

The net energy of solar electricity, now until 2050

I have been contemplating a post by Dr Simon Evans, deputy editor of Carbon Brief, called "Solar, wind and nuclear have ‘amazingly low’ carbon footprints, study finds".

The main point of this essay and the paper it refers to in Nature Energy (Pehl et al) is to argue that the carbon footprint of solar, wind and nuclear power looking forward to 2050 (using a partial life cycle analysis, that appears to exclude decommissioning of nuclear, but apparently includes the refining of nuclear fuel) is many times lower than coal or gas with carbon capture and storage. I don't have any disagreement with that, especially if rapid global decarbonisation proceeds in the next 32 years. (That is, if this proceeds, new solar and wind infrastructure will increasingly be fuelled by installed solar and wind.)

Energy return on energy investment and net energy

However the article in Carbon Brief also discusses energy return on energy investment (EROEI), though ambiguously this is called "energy return on investment” (EROI). [This is ambiguous as investment might be interpreted, by some, as financial investment.] EROIE is related to "net" energy - the higher the EROIE the higher the net energy.

The bottom line of this (leaving aside nuclear which I don't want to discuss here) is that the claimed EROEI (in 2050) for wind and solar is as high as 44:1 and 26:1, whereas the EROEI of coal is only 9:1. See the the figure in the Carbon Brief paper, reproduced from Pehl et al.

Note however that in the paper, it is the electricity return on energy investment which is actually calculated. I doubt that subtlety affects my main argument.

A 2016 paper by Louwen et al in Nature Communications, found "a break-even between the cumulative disadvantages and benefits of photovoltaics, for both energy use and greenhouse gas emissions, occurs between 1997 and 2018, depending on photovoltaic performance and model uncertainties."

I originally interpreted Louwen et al as meaning, under the most pessimistic assumptions, that an EROEI of one, for photovoltaic cells, was reached between 2005 and 2017, however, for cumulative energy production versus cumulative emissions. An EROEI of 1 is 26 times less than the Pehl et al paper (albeit for 2017 vs 2050). But on reading Louwen et al again, and after brief twitter discussion with Dr Simon Evans (and others) I now think that the EROEI for PV solar has increased by an extraordinary amount, and is probably over 20, at least for the best forms of solar, in reasonably sunny solutions. Such breakthroughs will not solve electricity generation in locations such as wintry Britain (see photo) but it is very encouraging. And, of course, affordable net energy does not guarantee sustainability, as energy cannot solve all aspects of limits to growth. But while affordable energy is not sufficient, it is necessary.

Solar panels in the snow: Reading, 2018 (photo Colin Butler)

Declining net energy and its risk to prosperity and even civilisation

I am by nature pessimistic, over the times scale of decades, and have published many articles warning that civilisation is heading towards a "check", which could be called Malthusian (eg "Sounding the Alarm: Health in the Anthropocene".) And I interpret a major cause for the global decline in wages growth for the non-elite to be both increased greed of the elite and declining net energy (as does Herman Daly, eg in "A further critique of growth economics Ecological Economics" 2013, 88, 20-24." See However, I am revising my view about the significance of net energy.

A new metric
A useful metric to develop is the ratio of EROEI for coal compared to solar (and wind). (Leave aside externalities such as air pollution, greenhouse gases and the pollution costs of obtaining rare elements needed for some or most renewable energy technology). Until I read Pehl et al, I would have guessed this ratio is about 40/2 = 20. It now be approaching 40/20 = 2; ie. much much better, and as coal declines and as solar improves the ratio will be even better.

I may further revise this post.