One of several incorrect and/or misleading criticisms of renewable energy in the Jeff Gibbs/Michael Moore film, Planet of the Humans, is the claim that building renewable energy technologies requires more life-cycle energy to be invested than they generate over their lifetimes.
If this notion were true, it would entail that the transition to renewables may displace investment in other important economic sectors.
However, this claim was only true for solar panels used to power artificial satellites several decades ago, before the panels were mass-produced; it was never true for commercial wind power.
It is generally agreed that the energy returned on energy invested (EROI) of once-through hydro-electricity is typically very large.
It is also widely accepted by energy experts that, in regions lacking large hydro-electric resources, renewable energy futures will be based on bulk energy from wind and/or solar, firmed up with storage, together with the electrification of most transport and non-electrical heat.
Already the majority of the respective annual electricity consumptions of Denmark, Scotland and South Australia is generated by wind and/or solar. Two North German states are operating on 100% netrenewable electricity, mostly wind.
The false myth about EROI of wind and solar photovoltaic (PV) is based on studies that use outdated data – one widely-cited reference even averaged EROIs over three decades. Such studies are invalid for three reasons:
- Wind and solar technologies have evolved rapidly, as reflected in the rapid decline in their costs.
- The cost of solar PV, in particular, has declined in real terms by about 90% over the past decade; wind by about 30%. Empirical data show that the energy invested in solar PV panels declined by an order of magnitude from 1997 to 2014.
- Another study has shown from empirical data how energy invested in solar PV panels decreases with the doubling of the cumulative capacity. Doubling continues at a high rate.
Recent evaluations find that EROIs of wind and solar located at suitable sites are generally much greater than one. Expressed another way, a large wind turbine generates the energy required to build itself in 3-6 months, while its lifetime is 25-30 years.
A solar PV panel generates the energy required to build itself in 1-2 years, while its lifetime is 20-25 years. Thus wind and solar PV can be considered to be breeders of RE.
That RE breeding is already occurring is witnessed by several reports in RenewEconomy that mining, mineral refining, steel-making and solar panel manufacturing are beginning to be powered by RE. Thus the life-cycle energy investments in RE will be provided by RE with zero greenhouse gas emissions.
Recent research finds that EROIs of fossil fuelled electricity technologies are comparable with EROIs of solar PV at medium-insolation sites and are less than EROIs of wind.
Therefore, the transition to energy systems based predominantly on renewable electricity could actually increase global EROI at the point of use.
Introducing storage complicates the issue. Some forms of storage are integrated with renewable electricity generation – e.g. hydro with dam; concentrated solar thermal with thermal storage; gas turbine burning a fuel such as ‘renewable’ ammonia produced from RE.
These technologies have EROIs greater than one, in some cases much greater. Other storage technologies are not net generators of electricity – e.g. pumped hydro, batteries; compressed air – and so reduce the EROI of the whole power system.
How much reduction is determined by the mix of storage types, the amount of storage required for reliability in the whole power system and the frequency of use of storage technologies.
Blakers, Lu and Stocks (2017) have shown by simulation modelling that the Australian National Electricity Market could be operated reliably with 92% of its energy supplied by variable renewables (wind + solar PV), existing hydro and a very small amount of energy storage capacity in the form of off-river pumped hydro.
This storage is used many times per year and so has high energy output. This result, together with the small quantity of storage required, suggests that the reduction in system EROI caused by storage would be small.
Similar results are likely for other regions of high wind and/or solar resource and low demand for winter heating.
In colder regions with low insolation, requirements for storage and/or additional transmission will be higher and the EROIs of systems involving power-to-gas and power-to-chemicals must be investigated.
So far, we’ve considered EROIs at a particular time, the present, and argued that wind and solar PV have sufficiently high EROIs to provide the bulk electricity for a zero-carbon energy future. Next the dynamic problem must be addressed.
Climate science demands a rapid transition to zero-carbon technologies. For simplicity, imagine that the zero-emission power stations are built and integrated into the grid in a series of rounds until no fossil fuelled power stations remain.
If the second round of new zero-emission power stations is constructed before the first round has begun to generate, then system EROI may indeed be temporarily less than one. However, when the rate of transition slows down, EROI will return to a value greater than unity.
This dynamic issue is not limited to RE – it applies to all rapid energy transitions introducing technologies with high up-front energy investments such as nuclear power and coal power with carbon capture and storage.
Although the latter two technologies generally have higher capacity factors (average power output divided by rated power) than wind and solar, they take much longer to build and so to generate an energy return to offset their high energy investments.
The longer lifetimes of a nuclear and coal-fired power stations are irrelevant to the temporary dynamic problem
If we accept the need for rapid, substantial climate mitigation, then we may have to accept a temporary decrease in system EROI during the period of rapid growth of RE. This is simply a recognition that we are facing a climate crisis.
These issues are discussed in much more detail in a paper currently under review in an international journal.
Mark Diesendorf is honorary Associate Professor in the Environment & Governance Group at the School of Humanities & Languages at UNSW Sydney