Energy Returns on Energy Invested revisited

Some say that renewable energy technologies do not generate much more energy than is needed for their construction and that their low energy return on energy invested (EROEI or EROI) will slow transition from fossil fuelled to renewable electricity. Indeed, some see the allegedly low EROIs compared with the historically high EROIs of fossil fuels as a major problem- with lower returns on investment, the global energy system will becomes less economically efficient.

A new study by Diesendorf and Wiedmann says that this view is based on outdated data on EROIs, on failing to consider the energy efficiency advantages of transitioning away from fuel combustion and on overestimates of the storage requirements for using variable renewable energy (VRE) sources. As an example of the impact of using dated data, it is pointed out that ‘Sers and Victor (2018) base their statement that EROIs of VRE technologies are ‘substantially lower than conventional fossil fuels’ on the meta-analysis of Hall et al. (2014), who ‘calculated the mean EROI value using data from 45 separate publications spanning several decades’. Averaging over several decades is invalid for solar PV and wind, because they have experienced huge improvements in technologies and supply chains, demonstrated by very large reductions in their respective prices as well as direct evidence’.

For the case of large-scale electricity supply, to try to do better, the paper draws upon insights from Net Energy Analysis and renewable energy engineering and shows that the EROIs of wind and solar photovoltaics, which it says can provide the vast majority of electricity and indeed of all energy in the future, are generally high (above 10) and rising. It adds that the impact of storage on EROI depends on the quantities and types of storage adopted and their operational strategies. In the regions considered in this paper, the quantity of storage required to maintain generation reliability is relatively small. So it is concluded that, in that context, overall, the transition to renewables may not be constrained by EROI considerations and that ‘contrary to traditional beliefs based on several previous studies, EROIs of wind and solar PV technologies at suitable locations are high and increasing’.

That view is shared, and developed more generally, in a paper by White and Kramer. They note that ‘many of the EROI values reported in the literature often refer to functionally non-equivalent energy carriers sampled at different stages of their respective supply chains, which has generated misleading comparisons.’ However, they say that, while ‘Energy return on investment (EROI) is a useful physical metric to compare the utility of energy production processes and their development over time, the concept has been extended from its physical, process-based origin to one that describes the societal metabolism. As such, EROI has been used to speculate about the prospects of humanity in a late-fossil and post-fossil civilization. Often, the narrative to emerge around EROI is that in the (near) future, prosperity will be compromised’.

By contrast they take a fresh look at EROI, ‘with a distinction between a physical EROI, which is useful at energy project level, and a societal, or economic, EROI—appropriate at the level of the whole economy. This distinction leads us to conclude that a renewable future is possible. Such a future is essentially unconstrained by the physical EROI and will have an acceptable economic EROI—not much different from that of the past century’.

As they see it, the EROI no longer matters as a society-wide measure ‘in the energy-abundant future that PV and wind energy technologies are about to open up for humanity’. So they say that the EROI concept needs rethinking. Basically there is no scarcity value to renewable sources, unlike finite fossil resources, although there will be access costs. However, these are falling fast, so are storage costs. So it does not matter too much if we need more raw primary energy input to get the same final output as we did from fossil fuels. In fact, the direct use of renewable energy avoids some of the energy conversion losses associated with the use of fossil fuel, as well of course as the pollution and carbon emissions. So the social and eco-costs are lower and the difference in net energy use may not be that great and may well reduce as EROIs improve further.

Interesting stuff, not least since it seems unlikely that the EROIs for nuclear technology, already low, will improve. Indeed, as high grade uranium ore becomes more scarce, more energy will needed to mine and process it, so the EROIs will fall even further. That’s not an issue that faces renewable like wind and solar- their operational input energy is freely available.

As I have pointed out in my new book, there may of course be some interim issues- while we build up the renewable energy conversion systems, we have to use fossil fuels to make the materials for that from ores (steel, aluminum etc), so there will be an emission debt. A role for carbon sequestration by forestry? But once established, the first wave of renewable generation plants can provide energy for building the next ones and the rest of the system, with no more emissions being produced.  And also, once the power system is all set up, there will be no need for new scare materials or more land-use, assuming we are heading for a steady state economy, just based on maintenance and renewal, not growth. A different world! But not an energy starved world.

The debate over EROIs can be rather tortuous, with varying assumptions being used about, amongst other things, the energy sources used for construction and making materials, and consequently quite a range of values being quoted. For, example, in a 2008 assessment based on 2002 data, Gagnon quotes the EROI for PV as being between 3-6, and for wind at around 18 for offshore and 34 for onshore, whereas a subsequent meta study by Harvey in 2010 puts the EROI for PV at between 8-25, and wind at between 40 and 80, depending on location. That may have been optimistic then, but may well now be the case, given technological advances since 2010.  In which case Diesendorf and Wiedmann, and also White and Kramer, may be right that, even given the need to deal with variability, as the technology continues improve, renewable EROIs will not be a major issue. After all, hydro, with reservoir storage,  is one of the key options for grid balancing, and, as all studies agree, hydro has a very good, EROI – 200 or more.