Dispelling myth of energy payback of renewable energy systems

It’s sometimes suggested that the life-cycle energy inputs of  renewable technologies can well exceed their outputs – an argument that some attempt to use against wind turbines, as well as solar PV modules. In this article by Mark Diesendorf  we find a clear explanation of just what ‘life-cycle energy inputs’ are, how they may be calculated, and how energy payback periods of various technologies compare. The figures may surprise you, and yet another myth bites the dust!

Published by Renew Economy. View original article 

One of the many myths propagating about renewable energy (RE) systems, especially solar photovoltaic electricity, is that the life-cycle energy inputs are greater than or comparable with the lifetime energy outputs. Is it generally true and, if not, under what special circumstances would it be reasonable?

Life-cycle energy inputs for solar PV modules are now greatly reduced
Life-cycle energy inputs for solar PV modules are now greatly reduced

Well, it makes for some purposes, such as industrial food production, in which energy inputs from the use of farm machinery and fertilisers are in total about ten times the somatic energy output from the food. Another example comes from the early use of solar photovoltaic (PV) modules to power human-made satellites.

In those days, before PVs were mass-produced, the energy input to manufacture them was indeed greater than their electrical energy output. At that time it made sense for that specialised purpose. However, nowadays solar modules are becoming an important energy source on Earth and they generate typically 10–25 times the energy required to manufacture them, as shown below.

It may also be reasonable to use a lot of low-grade energy in order to produce a little high-grade energy, provided the low-grade energy input is renewable and doesn’t require an expensive and energy-intensive collector system. An example is the generation of renewable electricity from the small difference in temperature between the ocean surface and deeper ocean produced by sunshine. However, in general it would be pointless to undertake coal mining, oil and gas recovery, nuclear power and any large-scale RE generation system in which life-cycle, high-grade, non-renewable, energy inputs were greater than the outputs.

Bluff Range wind farm under construction. Image source: Dave Clarke
Bluff Range wind farm under construction. Image source: Dave Clarke

By ‘life-cycle energy inputs’ I mean all the energy inputs from the mining, milling and processing of the raw materials, through the construction, operation and dismantling of the energy conversion system, to the management of the wastes. As the ‘easy’ reserves of fossil fuels and uranium are being used up, more ‘difficult’ sources are sought (eg, oil drilling in the deep ocean; mining low-grade uranium ore) and the energy inputs to these conventional energy systems are increasing.

Even more important in terms of environmental impacts is the source of the energy inputs, that is, whether it is fossil or renewable. If we wish to reduce CO2 emissions, there is little value in generating RE under circumstances where the emissions from fossil energy inputs are comparable with – or indeed greater than – the emissions saved by running the RE generator instead of a fossil fuelled generator. This is sometimes the situation with the production of biofuels.

A useful indicator of net energy generation of a technology is the energy return on investment (EROI), which is simply its lifetime energy output divided by the life-cycle energy inputs. Natural RE inputs – such as the solar energy used to grow biomass used to produce bioenergy – are not usually counted in the energy input. The bigger the EROI, the better. If the EROI is less than one, the technology is a net energy sink rather than a source.

Another useful indicator is the energy payback period, the number of years that an energy conversion system must operate in order to generate its total life-cycle energy inputs. It is equal to the lifetime of the technology in years multiplied by the ratio of energy input to energy output. Thus it is the lifetime divided by the EROI. The smaller the energy payback time, the better. For example, for a good quality solar PV system, the lifetime may be 25 years.

Then if the lifetime energy output equals 25 times the life-cycle energy input, the energy payback period is 1 year, which is being achieved now for some solar PV modules. On the other hand, if the energy output is equal to the energy input, then the energy payback period is equal to the lifetime of 25 years, which was the situation back in the 1980s when solar PV produced no net energy over its lifetime.

Actually, the calculation of the total energy input is not that simple, because decisions have to be made as to where to make a cut-off in what could be an infinite set of inputs. Do we include the energy required to make the machines that mined the raw materials that formed the energy conversion system of interest? A rigorous life-cycle analysis identifies a point where the additional inputs become negligible and cuts off the calculation there.

The whole process of calculating the life-cycle CO2 inputs becomes much easier as RE makes a bigger and bigger contribution to the national and global energy systems. Then RE conversion systems can be manufactured predominately by RE and they become RE breeders. Already some factories that make PV modules are powered either directly by rooftop PV modules or by purchases of ‘green’ electricity from a wind farm.

For the long-term there is an ambitious project proposal called the Sahara Solar Breeder in which factories would be built to manufacture silicon from the desert sand, solar PV cells from the silicon and solar power stations from the PV cells. Some of the electricity from the solar power stations would be used to expand the manufacturing of silicon and solar cells until enough electricity could be generated to supply all Africa and possibly the whole world.

The table gives approximate energy payback periods for some energy generation technologies. The range of values given for some technologies takes into account that energy outputs vary with location and fuel quality, if there is a fuel. For instance, a wind turbine at an excellent site may have an energy payback period of 3 months, but the same turbine at a moderate wind site may take 9 months to generate its energy inputs.

A silicon crystal PV module at an average US site may require 1.7 years while the same module in the sunny south-west USA may achieve 1.2 years. The short energy payback periods for wind and solar PV makes RE breeders a feasible proposition for rapid climate mitigation without producing a huge spike in GHG emissions.

Table 5.2 Energy payback periods for various technologies

Technology, Energy payback period (years), Expected lifetime of technology (years), Reference,
Solar PV modules  0.5–1.8  25  Fthenakis (2012) 
1.5 30 Raugei et al (2012)
Wind turbines (large) 0.25–0.75 20–25 Martinez et al. (2009)
Concentrated solar thermal (CST), parabolic trough 2 30 Desideri et al. (2013)
Nuclear (high-grade uranium) 6.5 30 Lenzen (2008)
Nuclear (low-grade uranium) 14 30 Lenzen (2008)
Coal: conventional 1–2 30 Raugei et al. (2012)
Coal with CCS ? ? ?

Notes: (a) Results for RE technologies vary sensitively with location.

(b) Some of the above results are contested in the scholarly literature on the basis of method.

This article is an edited extract from Mark Diesendorf’s new book, Sustainable Energy Solutions for Climate Change, published in Australia and New Zealand by UNSW Press in December 2013; to be published overseas by Earthscan in April 2014. The references are given in the book.

One thought on “Dispelling myth of energy payback of renewable energy systems

  1. I’m getting a bit long in the tooth now and my memory is not what it once was, but I recall some twenty years or so ago, a study was set up by the Australian Solar Energy Society to address this question.
    My recollection is that the result was that something like a few months was all that was needed for PV modules deployed in Australia to produce the equivalent energy required to make them.
    It would have been longer in locations with less radiant energy available.

    However, the picture showing several acres of PV modules used above in this post worries me a bit.
    How long will it be before the denialists start complaining about “infraglare” and set up a foundation to protect people from the ill effects of solar arrays?

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 )

Facebook photo

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

Connecting to %s