Understanding the divergence in EROI for solar PV
Some analyses calculate an EROI for solar PV of 60:1 or higher, while some calculate a much lower EROI, in one recent case, less than 1:1. The casual observer could be forgiven for wondering what’s going on. Since there is such a wide divergence, the default response seems to be to gravitate towards the results that reflect a individual’s worldview rather than trying to understand the role of solar PV in energy transitions.
Josh Floyd and I have recently published an analysis of solar PV.
Here is a shortened abstract:
Solar photovoltaics (PV) is widely regarded as one of the most promising renewable energy technologies. Results across studies can appear to diverge sharply, which leads to contestation of NEA’s relevance to energy transition feasibility assessment. This study explores how PV NEA approaches differ, including in relation to goal definitions, methodologies and boundaries of analysis. Here we show that most of the apparent divergence between studies is accounted for by six factors—life-cycle assessment methodology, age of the primary data, PV cell technology, the treatment of intermittency, equivalence of investment and output energy forms, and assumptions about real-world performance. The apparent divergence in findings between studies can often be traced back to different goal definitions. This study reviews the differing approaches and makes the case that NEA is important for assessing the role of PV in future energy systems, but that findings in the form of EROI or EPBT must be considered with specific reference to the details of the particular study context, and the research questions that it seeks to address. NEA findings in a particular context cannot definitively support general statements about EROI or EPBT of PV electricity in all contexts.
The aim of the study was to understand the reasons for the divergence rather than put a single figure in the conclusions. At one level, solar PV can be thought of as an efficient way to convert fossil fuels to electricity. In other words, investing fossil-fuelled electricity into solar plants will provide much more electricity over the PV lifetime. Within the context of our incumbent energy systems, the result is unambiguous – PV produces much more energy over its lifetime than is invested in its production.
The question arises as to how effective solar is at bootstrapping its own energy to increase the stock of PV, and substituting for the whole suite of electricity assets and energy services, including transport. In the long-run, there is much more uncertainty as to the value of PV in replacing incumbent infrastructure. Do read the paper.
Josh Floyd and I have recently published an analysis of solar PV.
Here is a shortened abstract:
Solar photovoltaics (PV) is widely regarded as one of the most promising renewable energy technologies. Results across studies can appear to diverge sharply, which leads to contestation of NEA’s relevance to energy transition feasibility assessment. This study explores how PV NEA approaches differ, including in relation to goal definitions, methodologies and boundaries of analysis. Here we show that most of the apparent divergence between studies is accounted for by six factors—life-cycle assessment methodology, age of the primary data, PV cell technology, the treatment of intermittency, equivalence of investment and output energy forms, and assumptions about real-world performance. The apparent divergence in findings between studies can often be traced back to different goal definitions. This study reviews the differing approaches and makes the case that NEA is important for assessing the role of PV in future energy systems, but that findings in the form of EROI or EPBT must be considered with specific reference to the details of the particular study context, and the research questions that it seeks to address. NEA findings in a particular context cannot definitively support general statements about EROI or EPBT of PV electricity in all contexts.
The aim of the study was to understand the reasons for the divergence rather than put a single figure in the conclusions. At one level, solar PV can be thought of as an efficient way to convert fossil fuels to electricity. In other words, investing fossil-fuelled electricity into solar plants will provide much more electricity over the PV lifetime. Within the context of our incumbent energy systems, the result is unambiguous – PV produces much more energy over its lifetime than is invested in its production.
The question arises as to how effective solar is at bootstrapping its own energy to increase the stock of PV, and substituting for the whole suite of electricity assets and energy services, including transport. In the long-run, there is much more uncertainty as to the value of PV in replacing incumbent infrastructure. Do read the paper.
[bctt tweet="Meaningful, tweetable quote."]
I appreciate that the divergence referred to in the title is a function of system definition. But eventually there will be a simpler, more easily defined system. I refer to a future clean energy world that will have of necessity become all-electric because all major sources of clean energy, solar, wind and nuclear, produce electrical energy as their useful output. With a single energy carrier one can then envisage a situation where some electricity is invested in energy generation which then produces the energy to run all electrified end-uses in industry, commerce, transport, households, etc. The problem reduces to determining just two energy quantities that characterize this system at steady state. How much electrical energy will that world need to generate its desired level of prosperity? And how much electrical energy must it invest annually to service and renew the generating systems. The sum of the two gives the total annual energy requirement. To give an example, the future world might need 350 EJ electrical energy to yield its present prosperity level. It may need a further 50 EJ to service and renew generating systems. In that case it must produce 400 EJ annually. It will be necessary to optimize returns and minimize that annual investment by determining an optimum technology mix. Of course this is a simplified and idealized description. There are huge problems at every turn, especially in achieving full electrification of end uses. But I hope the scenario helps give some insights into the divergent EROI issue.
Yes, an all electric system would simplify parts of the analysis, particularly factor 5 from table 4 (Investment and output energy form equivalence). But the issue of boundaries and other factors would still remain as sources of divergence. As to steady state – that’s the million dollar question. What is it and how do we get there? What investments are required in the transition period? What would global civilisation look like under the conditions of steady state?
Thanks Graham. Of course the ‘steady state’ is an idealized prospect. But the way I describe it using a single energy carrier product does I think help clarify and focus attention on the important issue for right now – how much of a low-carbon electricity output needs to be recycled into production? Take those two numbers, 350 and 50. The 350 EJ estimate for global electrical energy needs may be rough but is in line with a couple in the literature. The 50 EJ figure is pure conjecture on my part. Should it be 10 EJ or 500 EJ? That is, should it be around 0.3% or 150% of required output? Does anyone know? Would it also be applicable to today’s mixed fuel (fossil + sun/wind + fissile material etc.) economy? That’s really what I’m trying to address, from my rather remote point of view.
Should it be 10 EJ or 500 EJ?
Sounds like a good research project. Unfortunately, nearly all scenario analyses ignore this question and more-or-less assume that the energy intensity of investments for providing energy supply will be the same or better than we have now. For those of us who doubt this, this is an important question.